Australian Autism Alliance Submission to the consultation on NDIS access changes

NDIS Access must reflect real life – from assessment to support

2 October 2026

Contents

• 1. Introduction
• 2. Overall position
• 3. The three framing points that shape this submission
• 4. Responses to the consultation questions
• 5. Conclusion
• Attachment A: What Autistic Australians have told us and what the evidence shows about functional capacity, masking, burnout and permanence
• Attachment B: Implementation and accountability tests for the NDIS access reforms
• Attachment C: Autism specific assessment safeguards
• Reference list to Attachment A

1. Introduction

The Australian Autism Alliance (‘the Alliance’) welcomes the opportunity to respond to the Department of Health, Disability and Ageing (‘the Department’) consultation on NDIS access changes. This submission responds to each of the six consultation questions and sets out the safeguards the Alliance considers necessary if a functional capacity-based access process is to work fairly for Autistic people.

The Alliance is a funded national peak body and Disability Representative Organisation (DRO), providing a strong unifying voice for autism and authoritative policy expertise on autism. We work with Australian governments at all levels to strengthen policy and service systems affecting Autistic Australians. Our membership is a national network of organisations with diverse roles in autism, including organisations led by Autistic people, advocacy organisations, peak bodies, service providers and researchers. A number of our members work directly with Autistic people and their families through NDIS access, assessment, planning and review, bringing practical experience of how access settings operate in people’s lives every day.

With Autistic people comprising approximately 44 per cent of all active NDIS participants as at 31 March 2026, the design and implementation of the new access process are of profound importance to the autism community. The proposed replacement of the access lists with a standardised assessment of functional capacity would significantly change the access pathway for Autistic applicants and, through the proposed transition arrangements for reassessing existing participants, for current Autistic participants as well.

This submission has been informed by the diversity of perspectives, expertise and experience across the Alliance’s membership, including Autistic lived experience, advocacy, research. Through the Alliance’s Best Practice Policy Committee, these perspectives have informed the Alliance’s autism-specific and systems-level consideration of the proposed reforms across assessment, decision-making, transition, alternative supports, continuity and accountability. It also builds on the positions the Alliance put to the Senate Community Affairs Legislation Committee on the National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Bill 2026.

The submission also draws on the Alliance’s broader national evidence base, including the national consultation of more than 3,800 Autistic people, families and carers commissioned through the Australian Catholic University for the Senate Select Committee on Autism, as well as the NDIS Review, the Disability Royal Commission and the National Autism Strategy. It is also informed by Australian and international research on functional capacity, masking, Autistic burnout and permanence summarised at Attachment A. Attachment B offers practical implementation and accountability tests for the access reforms drawn from the Alliance’s Systems That Work™ approach.

In this submission, the Alliance uses Autistic people with very high support and dependency needs to include people sometimes described as having profound autism. Consistent with the National Autism Strategy, this includes Autistic people whose lifelong disability may involve 24-hour access to an adult for support or safety, substantial person-to-person assistance with daily living and personal care, complex communication needs, and/or significant ongoing monitoring.

The term recognises a population with very substantial and enduring support dependency; it is not intended to create a separate diagnostic gateway or to imply that every person has the same cognitive, communication or support profile1.

2. Overall position

The Alliance supports the objective of an access process that is clearer, fairer and more consistent. The NDIS Review identified that access decisions have depended too much on which diagnosis a person has, access to supporting evidence and on their ability to obtain the right paperwork. The Alliance members see every week how uneven the current evidentiary burden can be between people with comparable support needs. An access process that understands how a person’s impairment affects their everyday life is a legitimate objective. Whether the proposed process achieves it will depend entirely on its design.

Consistency should come from standardising the process, not standardising the person. Rights, safeguards, evidence standards and decision quality can and should be nationally consistent. But the application of that process must remain responsive to how disability is experienced in real life. Standardisation is a methodology, not an equity outcome, and a standardised process can still produce consistently inequitable decisions.

For Autistic people, this distinction is critical. Hence the Alliance’s support is conditional on how that principle is carried into practice. The consultation paper defines functional capacity as a person’s ability to undertake an activity in a standardised environment designed to minimise the impact of environmental and personal circumstances. Yet for Autistic people, environmental and personal circumstances can be integral to understanding Autistic functional capacity. Sensory load, communication demands, predictability, fatigue, masking, executive functioning and the presence or absence of trusted support change what an Autistic person can do from one setting to the next and from one week to the next. A process that measures capacity after removing that context will systematically understate disability for the largest cohort in the Scheme. Consistency should mean consistent rights, safeguards, evidence standards and decision quality, not necessarily identical assessment. Any standardised component should demonstrate, through appropriate validation and cohort specific testing, that it supports sound decision making without systematically under identifying particular forms of disability, and validity and equity should be demonstrated before an assessment methodology is relied upon for access decisions. It must be applied to evidence of how the person actually functions across settings and over time, with and without the supports currently holding their life together. The test the Alliance applies is simple: make the invisible visible, not require an Autistic person to perform their disability in an assessment encounter.

1 The National Autism Strategy 2025–2031 uses “Autistic people with very high support needs” and recognises a range of terminology including profound autism and high dependency needs. A 2026 consensus study has subsequently proposed a more specific research definition of profound autism based on markedly reduced adaptive functioning and need for adult supervision, together with severe cognitive impairment and/or very limited verbal communication. The authors caution that this remains a research definition rather than a clinical or service eligibility definition.

Reliance on an assessment methodology should not exceed the evidence supporting its validity. Where validity is established for the relevant population and circumstances, the methodology may be used accordingly. Where validity is uncertain or there is an elevated risk of false negatives, additional evidence or appropriately qualified professional judgement should be triggered.

The process must also work for Autistic people with very high support and dependency needs. In this submission, this refers broadly to Autistic people whose lifelong disability involves very substantial dependence on others for daily living, communication, health or safety, including people who require continuous or near-continuous supervision or support. The term is used to make this cohort visible, not to create a separate diagnostic gateway or assume that every person will have the same profile.

For this cohort, proportionality should generally reduce rather than increase assessment burden. Where existing evidence already clearly establishes lifelong impairment and very significant functional impact, access or reassessment should ordinarily be confirmed from that evidence. Additional assessment should be triggered only where a material evidentiary question relevant to the decision remains.

Two further conditions follow. First, permanence must be assessed as a question about the impairment, not about whether supports are working. The new process must keep four things distinct: the permanence of the impairment, the person’s functional capacity at a particular point in time, the person’s functional trajectory, and the support required to participate. Autism is lifelong. Therapy, education, adaptation and support can improve a person’s skills, participation and wellbeing without removing the underlying impairment. Capacity building is not cure, and effective support is not evidence that support is no longer required.

Second, where an access decision relies on another system being able to meet a person’s identified needs, there should be an identified and functioning receiving pathway. Where that pathway is refused, materially delayed, inaccessible, unsuitable or disputed, this should trigger rapid resolution and, where necessary to prevent a material gap in essential disability support, time-limited continuity while responsibility is resolved. A referral is not an outcome.

Proportionality should operate in both directions. Where evidence is uncertain or an assessment methodology may under-identify disability, additional safeguards should be triggered. Equally, where existing evidence already clearly establishes lifelong impairment and very significant functional impact, including for many Autistic people with very high support and dependency needs, the process should become simpler rather than more intensive. Government should not expend assessment capacity repeatedly proving what reliable evidence already establishes.

The Alliance’s participation in this consultation should not be read as endorsement of the removal of diagnosis based access, of the legislative changes already made through the NDIS Amendment (Securing the NDIS for Future Generations) Bill 2026, of the timetable for implementation, or of the broader reforms. The Alliance’s positions on that legislation remain as set out in our submissions to the Senate inquiry. Following passage of the legislation, the Alliance’s focus is on implementation: ensuring that the new access process is valid for Autistic people, that reliance on assessment methods reflects the evidence supporting their validity, and that people are not left without support while the systems around the NDIS develop.

Foreseeable risks in assessment, decision-making and transition are identified and addressed.

An access process cannot be judged in isolation from the systems that people found ineligible will be asked to rely upon. The Alliance applies an end-to-end systems test across six domains: assessment, decision making, transition, alternative supports, continuity and accountability, set out in section 3(c) and Attachment B. A technically sound assessment cannot compensate for an unavailable alternative service, an inaccessible evidence pathway, inadequate assessor capability, a lack of procedural fairness or the absence of continuity safeguards. The test should follow the person end-to-end: from assessment and evidence, through decision making, reasons and opportunities for correction or review and, where another pathway is identified, through supported referral and acceptance to actual access to appropriate support with continuity. It should provide a clear pathway for escalation and resolution when the system does not work, with monitoring capable of identifying and correcting systemic problems.

The Alliance has attached a short evidence summary drawing together the Australian and international research most relevant to the six consultation questions (Attachment A).

The Alliance’s five tests for fair access reform

Across the detailed recommendations in this submission, the Alliance applies five simple tests to the design and implementation of the new access process:

1. Assess functioning as it is lived. Functional capacity must reflect real-world functioning across settings and over time, including the environmental conditions, adjustments and supports that shape what a person can actually do.

2. Make the invisible visible. Assessment must make visible both disability that may not be apparent in an assessment encounter and the support that makes functioning possible. This includes masking, compensation, fluctuating capacity and the informal, environmental and funded scaffolding that can make supported functioning appear to be independent capacity.

3. Match reliance to validation. Assessment methodologies should be relied upon only to the extent that their validity has been demonstrated for the populations and circumstances in which they are used. Where validity is uncertain or false-negative risk is elevated, additional evidence or professional judgement should be triggered.

4. Separate permanence from support effectiveness. Improvement through treatment, therapy, education, adaptation or support does not make a lifelong impairment temporary or establish that support is no longer required.

5. Make transition real, and protect people when it fails. Where another system is relied upon, there must be an identified receiving pathway. Failed, delayed, inaccessible or unsuitable transitions should trigger rapid resolution and, where necessary, time-limited continuity to prevent a material gap in essential disability support.

Together, these tests support a proportionate approach: streamline where reliable evidence already establishes the answer; trigger additional protection where evidence, uncertainty or system failure creates material risk; and use monitoring to identify and correct recurring problems.

The Alliance’s detailed recommendations below translate these five tests into practical requirements for assessment, decision-making, transition, safeguards and accountability. They are intended to support consistent and efficient decision-making while concentrating additional process and safeguards where evidence or risk requires them.

Summary of recommendations

The Alliance’s recommendations are summarised here and developed through the paper.

Recommendation 1: Assess functional capacity as it is lived and make the invisible visible. Define and assess functional capacity as it is experienced in the person’s real environments, over time, and both with and without the informal, environmental and funded supports currently in place, consistent with the World Health Organization International Classification of Functioning, Disability and Health. The assessment should establish what the person can do, what enables them to do it, what it costs them to sustain it, and what happens if that support is removed or reduced. It should test whether an activity can be done safely, reliably, repeatedly and sustainably, including the preparation, prompting, support and recovery required. It should distinguish observable achievement, supported functioning and independent capacity; test whether activities can be performed safely, reliably, repeatedly and sustainably; and be capable of actively eliciting needs that are hidden, compensated for or normalised, while preserving the person’s voice, autonomy and supported decision making. A standardised assessment should be considered alongside the available evidence, with the weight placed upon it reflecting its demonstrated validity for the person’s relevant disability, communication, cognitive and support profile, and should never solely determine, an access decision. Where limitations or material uncertainty arise, decision makers must be able to consider the full range of clinical, functional, observational and longitudinal evidence, including evidence from families, carers, kin, educators and practitioners who know the person well, together with appropriately qualified professional judgement.

Recommendation 2: Strengthen the guiding principles so they protect against predictable inequity. Amend the draft guiding principles so that each becomes an operational and measurable design requirement, and so that they expressly provide that: functional capacity is dynamic and context dependent; a strengths based approach must not allow strengths in one area to offset substantial functional impairment in another, and intelligence, education, employment, verbal ability, adaptation or achievement are not proxies for low disability related support need. Diagnosis should neither automatically determine nor prejudice NDIS access, with relevant diagnostic and disability specific evidence remaining part of a multi source assessment of permanence and functional impact. The principles should also recognise reasonable adjustments as necessary to the validity of the assessment, not merely to its accessibility; neurodiversity-affirming practice, communication and sensory-safe, masking-aware, co-occurring disability, intersectionality and trauma-informed practice are named; and the importance of predictability (what is happening, what evidence is required, how it will be considered, what happens next and what the person can do if something goes wrong). Also add a principle on early intervention, continuity, lifelong outcomes and a genuine no wrong door safeguard.

Recommendation 3: Standardise the process, not the person by turning the guiding principles into measurable standards. Translate each guiding principle into nationally consistent and measurable standards for rights, safeguards, evidence requirements, assessor capability, assessment tools, workforce capability, communication, decision quality, review, system performance and accountability. Standardisation should create consistency in how decisions are made, not require people to demonstrate disability in the same way or under standardised conditions that remove the context necessary to understand their functional capacity. Any standardised component should demonstrate, through appropriate validation and cohort-specific testing, that it supports sound decision-making without systematically under-identifying particular forms of disability. Application must remain responsive to the person’s disability, communication, sensory and environmental context, culture, circumstances, existing formal and informal supports, and variability of functioning across settings and over time. Consistency should mean consistent rights, safeguards, evidence standards and decision quality - not identical assessment or identical application.

Recommendation 4: Keep permanence separate from whether support is working. Retain recognition of autism as a lifelong neurodevelopmental disability and distinguish four things distinct in legislation, rules and guidance: the permanence of impairment, functional capacity at a particular point in time, functional trajectory, and the support required to participate. Clarify that capacity building is not cure; that improvement in skills, participation or independence through treatment, therapy, support, education or adaptation does not make an impairment temporary or show that support is no longer required; that no one is required to undertake or exhaust treatment before permanence can be established; that treatment may be relied upon only where it is evidence based, clinically appropriate, available, accessible, affordable, safe and reasonable for that particular person; and that any adverse decision relying on treatment must identify the treatment, the evidence of likely benefit, its accessibility, risks and likely timeframe, and why it is reasonable for that individual.

Recommendation 5: Assess children on current evidence, not speculative future improvement. Ground access decisions for children in current evidence of disability and functional impact, considered alongside developmental trajectory, existing supports and foreseeable support needs, without speculative assumptions that future intervention will remove disability. Early intervention is intended to improve functioning, and improvement must not itself be used as a reason to question permanence of a lifelong impairment or significance of functional impact, decline access and redirect a child to Thriving Kids or other Foundational Supports. Where Thriving Kids or another Foundational Support is identified as the appropriate pathway, the transition should be supported and the receiving pathway identified. Where that pathway is unavailable, inaccessible, unsuitable, materially delayed or unable to meet the child’s identified needs, this should trigger escalation and resolution and, where necessary, protection against a material support gap.

Recommendation 6: Require demonstrated autism and disability-specific assessor capability. Require autism and disability specific capability for everyone who assesses functional capacity or makes access decisions, backed by national competency requirements, supervision, decision calibration, file audit and independent quality assurance, not simply completion of autism training. Competency should cover communication differences, sensory processing, masking and camouflaging, executive functioning, interoception, shutdown, demand related distress, Autistic burnout, co-occurring conditions and the variability of presentation across age, gender and culture.

Recommendation 7: Match reliance to validation Functional capacity assessment methods should only be relied upon to the extent that their validity has been demonstrated for the population and circumstances in which they are used. This is consistent with the Alliance’s recommendation to the Senate inquiry and the lessons of the abandoned 2021 independent assessments. Where validity has not been established, or there is an identified elevated risk of false negatives, the assessment should not be determinative and additional evidence or appropriately qualified professional judgement should be required. Validation should specifically examine false negatives, masking, context dependent functioning, assessment modality, the effect of existing formal and informal supports, very high support and dependency needs, and whether assessor capability materially changes outcomes, across Autistic people with different communication, cognitive and support profiles. Validation and outcome data should inform whether application is expanded, modified or restricted. The question is not only whether the tool is reliable, but who the tool gets wrong (Attachment C).

Recommendation 8: Make evidence requirements proportionate and do not reassess what is already reliably known. Provide proportionate pathways that do not turn on the visibility of disability: streamlined confirmation where permanence and significant functional impairment are already clearly established on the evidence, recognising that ‘clearly eligible’ must not become shorthand for visibly disabled; for current participants, reliance on the years of evidence about permanence, functional impact and support need that the Agency already holds, so that no one is required to prove again matters already well established unless there is a genuine evidentiary reason; reuse of existing evidence before new evidence is requested; and government funded assessment where the decision maker requires more than the applicant already holds, so that evidence requirements never become a capacity to pay test for NDIS access.

Recommendation 9: Design the access process around the person so people can navigate, challenge and resolve access decisions. Design the process around the person’s experience: a no wrong door entry point. Ensure independent disability advocacy is available where vulnerability, communication barriers, decision complexity, an adverse decision or a failed transition creates a heightened risk that the person cannot effectively navigate or challenge the process. This includes a plain language evidence plan that tells people what is happening, what evidence is required, how it will be considered, what happens next and what the person can do if something goes wrong;

accessible participation through interpreters, augmentative and alternative communication, Easy Read, visual supports, flexible modes and timing and supporter involvement at no cost; reasons that connect the evidence to each legislative criterion; a timely review pathway; and continuity of existing supports while a review or referral is unresolved. This is particularly important where support coordination and connection arrangements are reduced, delayed or unavailable (current rollout schedule is June 2028). Independent disability advocates should be recognised not only as supports for individuals, but as an early warning mechanism for systemic problems, including inaccessible assessment processes, failed referrals, service gaps and recurring transition failures. Government should establish a mechanism through which these patterns can be escalated and inform continuous improvement.

Recommendation 10: Make the process work across disability, cultural and geographic contexts. Co-design culturally safe, community controlled and place based pathways with First Nations people and culturally and linguistically diverse communities; not requiring specialist evidence that is unavailable locally in regional, rural and remote areas; treating family, carer, kin and supporter evidence as valid evidence while preserving the person’s voice, consent and supported decision making rights; and never equating verbal ability, eye contact, academic achievement or short term task completion with low support need.

Recommendation 11: Build safe transition and trigger protection when transition fails. Where another system is identified as appropriate, there should be an identified receiving pathway capable of meeting the person’s relevant needs, with clear responsibility for referral, reception and resolution. A referral is not an outcome: an effective transition requires referral, acceptance and timely access to appropriate support. Where a referral is refused, materially delayed, inaccessible, unsuitable, fails to provide the identified support, or responsibility between systems is disputed, this should trigger rapid escalation and resolution and, where necessary to prevent a material gap in essential disability support, time-limited continuity while responsibility is resolved. No person should be required to resolve a disagreement between systems themselves. Governments should publish sufficient information about the availability, eligibility, capacity and wait times of the pathways being relied upon so that readiness and emerging gaps can be assessed. Accountability for the outcome should rest with an identified government decision-maker rather than be diffused across systems, with it possible to answer at each interface who is responsible, for what, from when, until when, what happens when the pathway fails and who owns escalation and continuity. Governments should monitor failed referrals, wait times and unresolved transitions so that recurring failures trigger system-level correction rather than repeated individual exceptions.

Recommendation 12: Build co-design, transparency and correction into implementation Involve Autistic people, families and autism representative organisations in the design, testing, implementation and evaluation of the new process, including advice to the Technical Advisory Group and lived experience governance. Publish disaggregated access and review outcomes, including review and overturn rates, by disability cohort, age, gender, geography and cultural background, with independent evaluation. Set predefined indicators for

identifying systemic inequity and equity and quality thresholds with authority to pause, modify or withdraw tools and decision rules that produce systematic inequity. Publish a cumulative impact assessment of the access reforms alongside the other NDIS changes commencing between 2026 and 2028.

A proportionate safeguard framework

The Alliance does not propose that every safeguard should add process to every access decision. Additional protections should be triggered where identifiable evidence, uncertainty or system failure creates a material risk of an incorrect decision or support gap.

This approach concentrates safeguards where risk is identified rather than adding additional process to every access decision. It also prioritises reuse of existing evidence, proportionate assessment and early resolution of failed pathways. The purpose is to prevent avoidable reassessment, review, appeal, service duplication and escalation while protecting people from incorrect access decisions and material gaps in essential disability support.

| advocacy concerns | Investigate and correct the relevant assessment method, decision rule, workforce practice or pathway, with authority to modify, restrict or pause the affected element where predefined thresholds are exceeded. | | Very high and established support dependency | Existing evidence establishes lifelong impairment and very significant functional impact, including a sustained need for very substantial assistance or supervision | Use streamlined confirmation rather than full reassessment. Reuse existing Agency and other reliable evidence. Require additional assessment only where a material evidentiary question remains relevant to access. |

Implementation and accountability

Each safeguard should have a defined trigger, responsible decision-maker and required response.

Assessment and access decision-makers should record applicable triggers as part of the decision record, while transition failures should also be capable of being raised by the person, their supporter, independent disability advocate or receiving service.

For the very-high-support-dependency trigger, the decision-maker should record the evidence establishing that the streamlined pathway applies, rather than requiring the person to establish a new diagnostic category or undertake an additional assessment solely to qualify for streamlining.

Systemic triggers should be identified through aggregated access, review, transition, complaints and advocacy data. Trigger events, responses, resolution timeframes and outcomes should be recorded consistently and reported publicly in aggregate and by disability cohort.

Government should regularly report on trigger frequency, resolution, review and overturn rates, failed transitions and recurring disparities, using predefined thresholds to identify where assessment methods, decision rules or service pathways require correction.

3. The three framing points that shape this submission

a. Make the invisible visible: functional capacity must be assessed as it is lived: across settings, over time and with the supports that sustain it

The consultation paper’s definition of functional capacity, a person’s ability to undertake an activity in a standardised environment designed to minimise the impact of environmental and personal circumstances, is the single most important design choice in the reform. It also creates a particular risk for Autistic people. The contemporary international standard, the World Health Organization International Classification of Functioning, Disability and Health (ICF), treats environmental factors as part of the classification of functioning itself: disability arises in the interaction between an impairment and the physical, social and attitudinal environment in which the person lives. An assessment that removes environmental and personal context may therefore remove information necessary to understand disability as it is actually experienced. The overarching requirement for Autistic people is that the process makes the invisible visible, rather than requiring a person to perform their disability in an assessment encounter.

For Autistic people, the interaction between person and environment is decisive, but it is not random. Members who assess and support Autistic people every day can predict, describe and document how sensory load, communication demands, unpredictability, executive functioning demands, fatigue and the availability of trusted support change a person’s functioning. Many Autistic people perform capably in a structured, quiet, one to one setting and cannot sustain the same performance in a classroom, a workplace, a shopping centre or a busy home. Masking and camouflaging, which the research shows to be effortful, costly to mental health and more common among Autistic women, mean that what an assessor observes in a short encounter is often the product of enormous hidden effort rather than a picture of everyday capacity. Autistic burnout, now well described in Australian research, involves the loss of previously held skills and reduced tolerance to demands, so capacity measured on one day is a moving target. Autistic functional impact may be context dependent and variable; masked or compensated for; affected by sensory and communication environments; dependent upon predictability, routine and trusted people; affected by executive functioning, interoception, shutdown, burnout and demand related distress; different across settings and over time; and compounded by co-occurring disability and health conditions. The process must therefore assess sustainable real world functioning, not whether someone can perform an activity once in a structured setting. The relevant question is whether something can be done safely, reliably, repeatedly and sustainably, including what preparation, prompting, support and recovery are required.

Two consequences follow for design.

First, existing supports conceal underlying need. Family support, prompting, supervision, routines, technology, school and workplace adjustments, environmental modifications and funded supports can all create the appearance of capacity. The process should make the support around the person visible and assess functioning both with and without it, so that it understands what the person can do, what enables them to do it, what it costs them to sustain it, and what happens if that support is removed or reduced. Effective support must not become evidence that disability or support need has disappeared. Three things must be kept distinct: observable achievement, supported functioning and independent capacity. Completing an activity, attending school or work, communicating effectively or participating in the community does not establish that the person can do so independently, and observable achievement should not be treated as evidence of independent functional capacity without identifying the support, adjustments, effort and recovery required to achieve it.

The methodology must also be capable of actively eliciting needs that have become hidden, compensated for or normalised over time. Some Autistic people under report functional difficulty because of masking, literal interpretation of questions, acquiescence, difficulty identifying internal states, learned adaptation, or because substantial family or other support has become their normal baseline. An assessment that depends on a person independently identifying and articulating every area of need will miss them, and the characteristics of a person’s disability will themselves have become a barrier to demonstrating disability related functional impact. Second, a standardised tool can only ever be one input. Consistency should mean consistent rights, safeguards, evidence standards and decision quality, not necessarily identical assessment, and never a single tool, a single interview or a single day. The Alliance recommends a structured professional judgement model in which national standards and decision rules are applied to multiple sources of evidence, including diagnostic and specialist reports, functional assessments, observation across settings, school and employment information, family, carer and kin accounts, the person’s own account and support records over time.

Proportionality should operate in both directions. This does not require every person to undergo a more intensive assessment. Existing reliable evidence should be reused wherever possible. Where masking, fluctuating or context-dependent functioning, communication differences, conflicting evidence or limitations in the assessment method create a material risk of under-identifying disability, that risk should trigger additional evidence or appropriately qualified professional judgement. Equally, where reliable evidence already establishes the relevant matters, further assessment should not be required simply for the sake of standardisation.

At the other end of the assessment problem are Autistic people with very high support and dependency needs, including people described as having profound autism, whose lifelong impairment and very significant functional impact may already be extensively documented. For this cohort, proportionality should ordinarily mean confirmation rather than re-assessment. The process should not require direct performance of activities or repeated functional testing merely to reconfirm established dependence. Reliable longitudinal evidence, including from families, carers, communication partners and support providers, may provide a more valid account of functioning than requiring the person to demonstrate incapacity in an unfamiliar assessment encounter.

Australia has already tested the alternative. Independent functional assessments using standardised tools were abandoned in July 2021, and the Joint Standing Committee on the NDIS found that standardised assessment tools may not be appropriate for certain cohorts of people with disability and that the tools lacked relevance, sensitivity and specificity. The international consensus work on ICF core sets for autism concluded that no standardised, internationally accepted tools exist to assess autism related functioning. The Department’s own paper acknowledges that assessment approaches, tools and thresholds are yet to be developed.

The Alliance therefore asks that the lesson of 2021 be built into the design from the outset: publish the methodology, independently validate it for its intended purpose and for the populations and circumstances in which it will be used, including Autistic people and other cohorts, test for false negatives and not only for consistency between assessors, and demonstrate workforce capability. The degree of reliance placed on an assessment should reflect the evidence supporting its validity for the population and circumstances in which it is used. Where validity is uncertain or there is an elevated risk of false negatives, additional evidence or appropriately qualified professional judgement should be triggered. The Alliance asked this question of the Bill in May 2026 and it remains unanswered: how will assessors distinguish between sustainable functioning and functioning maintained through masking, exhaustion or informal family scaffolding? The Alliance’s prior work on this question is summarised at Attachment A, section 10, and the autism specific safeguards it supports are set out at Attachment C.

b. Keep permanence separate from whether support is working.

Autism is a lifelong neurodevelopmental disability. The international diagnostic classifications, the World Health Organization and the National Institute for Health and Care Excellence all describe it that way, and the NDIA’s own access lists treat a Level 2 or 3 autism diagnosis as likely to meet the disability requirements. What changes over a lifetime is not whether a person is Autistic but how their impairment interacts with the demands and supports around them. Four things must therefore be kept distinct: the permanence of the impairment, functional capacity at a particular point in time, functional trajectory, and the support required to participate.

The consultation paper proposes clearer guidance on how treatment, intervention and rehabilitation should be considered when assessing permanence. The Alliance supports clarity, and asks that the guidance make three things explicit. Improvement in skills, communication, confidence, participation or independence achieved through therapy, education, adaptation or support is evidence that the support is working and necessary; it is not evidence that the impairment is temporary or that support is no longer required.

The existence of a theoretical treatment does not mean a person must undertake it before permanence can be established: treatment relied upon in an access decision should be evidence based, clinically appropriate, available, accessible, affordable, safe and reasonable for that particular person, and no one should be required to undergo invasive, risky, repetitive or clinically inappropriate treatment, or to exhaust every intervention, before permanence is accepted. And where a decision maker proposes to rely on the proposition that treatment is likely to improve functional impact, the decision must identify the treatment, the evidence of likely benefit, its accessibility, its risks and likely timeframe, and explain why it is reasonable for that individual.

This matters most for children. Autistic children’s functional impact and support needs frequently change, and often become more apparent, as developmental, emotional, educational and social demands increase through primary school, adolescence and the transition to adulthood. The Alliance is particularly concerned about circular reasoning: a child improves with intervention, permanence or significance is questioned, NDIS access is declined, and the child is directed to Thriving Kids or Foundational Supports. Early intervention is intended to improve function. Improvement does not mean a lifelong impairment is temporary, and the Australian evidence on early intervention shows that the cost of getting these decisions wrong is paid in developmental windows that do not reopen. Equally, a snapshot of a child’s current functioning is not enough. Decisions should be grounded in current evidence of disability and functional impact, considered alongside developmental trajectory, existing supports and foreseeable support needs, without speculative assumptions that future intervention will remove disability.

At the same time, not every Autistic child will require the NDIS, and the existence of Thriving Kids or another Foundational Support should not prevent a child being directed to that pathway where it is appropriate to their needs. Where an access decision relies on that pathway, however, and the proposed alternative is unavailable, inaccessible, unsuitable, materially delayed or unable to meet the child’s identified needs, this should trigger escalation and resolution and, where necessary to prevent a material gap in essential disability support, time-limited continuity while responsibility is resolved. The objective should be neither automatic NDIS continuation nor automatic redirection, but the right pathway based on current evidence, with a safeguard when that pathway cannot deliver what the decision assumes it will provide.

c. Readiness before reliance: validate the assessment, establish the pathway and protect people when it fails

The reforms ask people to rely on two things that do not yet exist: a validated functional capacity assessment, and a set of Foundational Supports and mainstream responses capable of meeting the needs of people who are found not to require the NDIS. Thriving Kids commences from October 2026 with full rollout by January 2028, the same date from which access reforms apply. Later phases of Foundational Supports are not yet designed. The Technical Advisory Group is yet to advise on assessment approaches, tools or thresholds. The Alliance put to the Senate inquiry that the safer sequence is to design, test and build the new arrangements and

demonstrate their readiness before people are transitioned onto them, and that implementation risk should sit with government rather than with the person whose access is changing. That underlying principle remains: the particular assessment method or alternative pathway being relied upon should be sufficiently valid and operational for the people to whom it is being applied, with defined safeguards where that assumption proves wrong.

A referral is not an outcome. Where a person is directed to another service system to rely on, an effective pathway requires referral, acceptance, timely access to appropriate support and continuity until responsibility has genuinely transferred. The existence of an alternative service or system should not be relied upon in an NDIS access decision unless that pathway is funded, available, accessible and capable of meeting the person’s identified needs. Where a referral is refused, materially delayed, inaccessible, unsuitable, fails to provide the identified support, or responsibility between systems is disputed, this should trigger rapid escalation and resolution and, where necessary to prevent a material gap in essential disability support, time-limited continuity while responsibility is resolved.

Governments publish the evidence of readiness, including regional service maps, capacity, wait times and accountability for gaps, before access boundaries change so that the capacity of the pathways being relied upon can be assessed and emerging gaps identified.

The Alliance’s Systems That Work™ readiness and accountability tests (Attachment B) offer a practical way to do this. Accountability for whether a particular person is supported should rest with an identified government decision maker. A framing in which the NDIS, Foundational Supports, mainstream services and community supports are described together as an ecosystem must not be allowed to diffuse that accountability across systems, because a person who falls between them has no one to hold responsible. Systems thinking should sharpen accountability, not disperse it. At each interface it should be possible to answer: who is responsible, for what, from when, until when, what happens when the intended pathway fails, and who owns escalation and continuity. That is the difference between describing an interconnected service ecosystem and applying a Systems That Work™ test to it: the test requires responsibility, readiness, triggers and accountability to be identifiable at each point in the pathway.

The same principle should operate at system level. Governments should monitor failed referrals, transition timeframes, access and review outcomes, complaints and advocacy signals so that recurring failures or disparities can be identified. Where predefined indicators show that an assessment method, decision process or service pathway is producing recurring problems, this should trigger investigation and proportionate corrective action. The Alliance also asks that the Department be candid about scale. Reassessing every current participant over three years, with Autistic people comprising around 44 per cent of the Scheme, is a workforce and quality undertaking of a different order from anything the NDIA has attempted. The Commonwealth Ombudsman has previously found that NDIA review delays pose a particular risk to people who may lose services or experience deterioration. For people already in the Scheme, the Agency holds years of evidence about permanence, functional impact and support need, and reassessment should not require people to prove again matters already well established unless there is a genuine evidentiary reason. The proportionate pathways recommended in this submission, including streamlined confirmation where permanence and significant functional impairment are already clearly established, are as much a matter of system feasibility as of fairness, provided that ‘clearly eligible’ does not become shorthand for visibly disabled.

Recommendation 1 (headline recommendation): Define and assess functional capacity as it is experienced in the person’s real environments, over time, and both with and without the informal, environmental and funded supports currently in place, consistent with the World Health Organization International Classification of Functioning, Disability and Health. The assessment should establish what the person can do, what enables them to do it, what it costs them to sustain it, and what happens if that support is removed or reduced. It should test whether an activity can be done safely, reliably, repeatedly and sustainably, including the preparation, prompting, support and recovery required. It should distinguish observable achievement, supported functioning and independent capacity; test whether activities can be performed safely, reliably, repeatedly and sustainably; and be capable of actively eliciting needs that are hidden, compensated for or normalised, while preserving the person’s voice, autonomy and supported decision making. A standardised assessment should be considered alongside the available evidence, with the weight placed upon it reflecting its demonstrated validity for the person’s relevant disability, communication, cognitive and support profile, and should never solely determine, an access decision. Where limitations or material uncertainty arise, decision makers must be able to consider the full range of clinical, functional, observational and longitudinal evidence, including evidence from families, carers, kin, educators and practitioners who know the person well, together with appropriately qualified professional judgement

4. Responses to the consultation questions

The Alliance’s responses follow the structure of the consultation paper. The table below maps each of the Department’s six questions to the relevant sections, recommendations and attachments of this submission.

a. Do the draft guiding principles provide the right foundation for a future NDIS access process?

Broadly, yes. The seven draft principles recognise human rights, lived experience, evidence, person centred and strengths-based practice, equity, accessibility and continuous improvement, and the Alliance and its members support each of them. Members particularly welcome the commitment to a process that is proportionate, avoids unnecessary repetition and is intended to work for people whose disability is complex, fluctuating, episodic or not immediately visible, and the recognition that consistent outcomes are required regardless of disability type.

The principles will only provide a foundation, however, if they operate as enforceable design requirements rather than aspirations. As drafted they do not say how the tension at the heart of the reform, between national consistency and standardisation on one hand and person centred, context sensitive assessment on the other, will be resolved. The Alliance would resolve it this way: consistency should mean consistent rights, safeguards, evidence standards and decision quality, not necessarily identical assessment, and standardisation is a methodology, not an equity outcome. Nor do they yet say anything about the single feature of the proposed model that most concerns the autism community: that functional capacity is to be assessed in a standardised environment designed to minimise the very circumstances in which Autistic people’s functional impact occurs. The Alliance’s responses to questions 2 and 3 address both gaps.

The Alliance also asks that the principles be translated into measurable standards for assessment design, workforce capability, communication, decision quality, review and system performance, and that the Department publish, for any tool or method adopted, the validation evidence, the cohorts in which it has been tested and its known limitations and error rates. Trust in the reform will depend on the Department demonstrating that the principles are applied consistently across all disability cohorts and are not shaped by assumptions or cost objectives associated with particular diagnoses. A principle that cannot be tested against what actually happens to applicants is not a foundation.

The principles also need to be applied to the reform as a system, not only to the assessment methodology. A technically sound assessment cannot compensate for an unavailable alternative service, an inaccessible evidence pathway, inadequate assessor capability, a lack of procedural fairness, a failed transition to another service system or the absence of continuity safeguards. The Alliance therefore proposes an end to end systems test that follows the person from assessment and evidence, through decision making, reasons and opportunities for correction or

review, and, where another pathway is identified, through supported referral and acceptance to actual access to appropriate support, continuity, monitoring and the correction of systemic problems (section 3(c) and Attachment B). Critically, the existence of another system does not establish its readiness: Where an access decision relies on Thriving Kids, Foundational Supports or a mainstream system, there should be an identified receiving pathway appropriate to the person’s needs, with defined escalation where that pathway fails. Predefined indicators should identify recurring failures or systemic inequity and trigger investigation and corrective action. This allows implementation to proceed while ensuring that foreseeable failures are detected and addressed rather than borne by the individual.

b. Should any principles be added, removed or changed?

The Alliance recommends retaining all seven principles, converting each into an operational and measurable design requirement, and making the following changes.

i. Principle 1 (aligned to the NDIS Act and human rights) should translate into concrete entitlements: supported decision making, reasonable adjustments, communication access, independent advocacy, procedural fairness and accessible review rights. Reasonable adjustments are not simply an accessibility consideration: they can be fundamental to the validity of the assessment itself. Where communication, sensory, cognitive or processing differences affect how a person can participate, an assessment conducted without the necessary adjustments may not provide a valid basis for drawing conclusions about functional capacity. That proposition should sit across the human rights, evidence and equitable, inclusive and safe principles. A human rights principle that does not name the mechanisms through which rights are exercised in an access process will not change the experience of a single applicant.

ii. Principle 2 (informed by lived experience) should mean genuine co-design, testing and ongoing evaluation with diverse people with disability, including Autistic people with different communication, cognitive and support profiles, people with very high support and dependency needs, people with complex communication needs and people experiencing intersectional disadvantage, and should commit to disability specific expertise, including autism expertise, in design, implementation and evaluation. Reforms affecting Autistic people should be designed with the autism community, not for it.

iii. Principle 3 (evidence based and consistent) should state that consistency does not mean uniformity: consistency should mean consistent rights, safeguards, evidence standards and decision quality, not necessarily identical assessment. Any standardised tool must be validated across relevant disability cohorts, including Autistic people, with examination of false negatives and systematic bias, and with the validation evidence, known limitations and error rates published. Where validity is uncertain or an elevated false-negative risk is identified, additional evidence or appropriately qualified professional judgement should be triggered. The principle should also recognise that an assessment conducted without the reasonable adjustments a person needs is not valid evidence of their functional capacity. Consistency between assessors who are all measuring the wrong thing is not fairness.

iv. Principle 4 (person centred and strengths based) needs an explicit safeguard. A strengths based approach must not allow strengths in one area to offset substantial functional impairment in another. Intelligence, education, employment, verbal ability, adaptation or achievement should not be treated as proxies for low disability related

support need, and strengths, coping strategies and acquired skills must not be used to offset substantial functional impairment or support need. The principle should also state that functional capacity is dynamic and context dependent. Members support strengths based practice and are equally clear that strengths based practice is not the same as strengths being weighed in an eligibility determination.

• vi. Principle 5 (equitable, inclusive and safe) should expressly encompass neurodiversity affirming practice, communication and sensory safety, masking, co-occurring disability, intersectionality and trauma informed practice, and should name cultural safety for First Nations peoples and accessibility for culturally and linguistically diverse communities, applying before, during and after assessment, with reasonable adjustments treated as a condition of assessment validity rather than an optional accommodation.
• vii. Principle 6 (transparent, proportionate and accessible) should also include predictability. People should understand what is happening, what evidence is required, how it will be considered, what happens next and what they can do if something goes wrong. It should guarantee interpreters, augmentative and alternative communication, advocates, supporters and flexible assessment modes without cost to the applicant, require accessible reasons that connect the evidence to each legislative criterion, and provide a streamlined pathway where the evidence already clearly establishes permanence and significant functional impact.

Proportionality should operate in both directions: additional evidence where material uncertainty exists, and streamlined confirmation where reliable evidence already establishes permanence and significant functional impact.

A standardised component may form part of the evidence, with the weight placed upon it reflecting its demonstrated validity for the relevant population and circumstances. Where limitations or material uncertainty arise, decision-makers should consider broader clinical, functional, observational and longitudinal evidence and, where necessary, appropriately qualified professional judgement.

• viii. Principle 7 (continuous quality improvement) should require published, disaggregated outcomes, including review and overturn rates, and independent evaluation, together with lived experience governance and an ability to pause or change an assessment approach if it produces systematic inequity. This is the principle that will tell governments, and the community, whether the other six are working.
• ix. Add a principle on early intervention, continuity and lifelong outcomes: access arrangements should support timely intervention, prevent avoidable deterioration and recognise lifelong disability. Where movement between systems creates a material risk of delayed, reduced or interrupted essential disability support, this should trigger an identified escalation and resolution pathway and, where necessary, time-limited continuity while responsibility is resolved. The continuity safeguard is essential. Without it, an early intervention principle could be read as reinforcing the proposition that children should be moved from the NDIS to Thriving Kids because intervention may improve their functioning, which is the opposite of what the evidence on early intervention, summarised at Attachment A, supports. The principle should incorporate a genuine no wrong door safeguard so that people are not required to resolve system boundaries themselves.
• x. State the place of diagnosis expressly. Diagnosis should neither automatically determine nor prejudice NDIS access. Relevant diagnostic and disability specific evidence should

remain part of a multi source assessment of permanence and functional impact, because diagnosis is relevant evidence about the nature, permanence and expected course of an impairment and informs how functional evidence should be assessed. The reform should not move from over reliance on diagnosis to under reliance on it. No disability cohort should be advantaged or disadvantaged by the design of assessment tools, evidence requirements or eligibility thresholds, and decision making should be free of assumptions, stereotypes, cost saving objectives or stigma associated with particular cohorts. Members ask that the principles also govern how the reforms are communicated, so that public communication does not position Autistic people, or any cohort, as a policy problem or a cost driver. The need is not abstract: in a qualitative community survey conducted by Alliance member Amaze in July and August 2026, respondents described recent public commentary about autism and the NDIS as ‘dismissive’, ‘hurtful’ and ‘utterly unfair’, linked it to increased stigma, and reported feeling less safe disclosing that they are Autistic, with one respondent observing that ‘telling people I’m autistic feels the same as coming out in many ways, I can only do it if it’s safe’.33 A reform whose public framing deters people from disclosing their disability will not produce fairer access decisions.

Recommendation 2: Amend the guiding principles so that each becomes an operational and measurable design requirement, and so that they expressly provide that: functional capacity is dynamic and context dependent; a strengths based approach must not allow strengths in one area to offset substantial functional impairment in another, and intelligence, education, employment, verbal ability, adaptation or achievement are not proxies for low disability related support need; diagnosis should neither automatically determine nor prejudice NDIS access, with relevant diagnostic and disability specific evidence remaining part of a multi-source assessment of permanence and functional impact; reasonable adjustments are treated as necessary to the validity of the assessment, not merely to its accessibility; neurodiversity-affirming practice, communication and sensory-safe, masking-aware, co-occurring disability, intersectionality and trauma-informed practice are named; and the importance of predictability (what is happening, what evidence is required, how it will be considered, what happens next and what the person can do if something goes wrong). Also add a principle on early intervention, continuity, lifelong outcomes and a genuine no wrong door safeguard.

Recommendation 3: Translate each guiding principle into nationally consistent and measurable standards for rights, safeguards, evidence requirements, assessor capability, assessment tools, workforce capability, communication, decision quality, review, system performance and accountability. Standardisation should create consistency in how decisions are made, not require people to demonstrate disability in the same way or under standardised conditions that remove the context necessary to understand their functional capacity. Any standardised component should demonstrate, through appropriate validation and cohort-specific testing, that it supports sound decision-making without systematically under-identifying particular forms of disability. Application must remain responsive to the person’s disability, communication, sensory and environmental context, culture, circumstances, existing formal and informal supports, and variability of functioning across settings and over time. Consistency should mean consistent rights, safeguards, evidence standards and decision quality - not identical assessment or identical application.

c. What do we need to consider when designing an access decision-making process that is fair, consistent, transparent, disability-informed and evidence-based?

The Alliance’s central advice is set out in section 3(a): the future process should be built around a structured professional judgement model that combines national evidence standards and decision rules with qualified judgement informed by multiple sources of evidence, so that diagnosis based access is not simply replaced by tool-based access. The overarching requirement is that the process makes the invisible visible rather than requiring an Autistic person to perform their disability in an assessment encounter. The following design considerations give effect to that model.

• i. Assess real world functioning. Examine communication, self care, self management, mobility, learning, social interaction, community participation and participation in education or employment across the settings that matter to the person, and record environmental demands, available adjustments, support intensity, prompting, supervision, preparation and recovery time. For Autistic people, assessment must distinguish between the permanence of autism as a lifelong neurodevelopmental disability, the person’s functional capacity at a particular point in time and in particular contexts, and the supports required to participate in everyday life.
• ii. Assess sustainable performance over time. Use longitudinal evidence and the person’s overall pattern of functioning, and ask whether an activity can be done safely, reliably, repeatedly and sustainably, including what preparation, prompting, support and recovery are required, rather than whether it can be performed once in a structured environment. A single good day, a short appointment or a highly structured encounter should not outweigh persistent evidence from everyday life. Autistic burnout, in which skills previously held are lost and tolerance to demands reduces, arises from cumulative load without any new impairment, and is frequently misread as psychiatric illness; a process that cannot see it will refuse people at the point of greatest need.
• iii. Make the support around the person visible. Family support, prompting, supervision, routines, technology, school and workplace adjustments, environmental modifications and funded supports can create the appearance of capacity. The assessment needs to understand what the person can do, what enables them to do it, what it costs them to sustain it, and what happens if that support is removed or reduced, and should consider functioning both with and without current supports. Sustained participation is often achieved through substantial effort, accommodation and masking, at considerable personal cost, and can end in burnout, reduced capacity and increased support need. Effective support is evidence of need, not of its absence.
• iv. Distinguish observable achievement, supported functioning and independent capacity. Completing an activity, attending school or work, communicating effectively or participating in the community does not establish that the person can do so independently. Assessment should identify the prompting, supervision, co-regulation, preparation, routines, environmental modifications and human support that make the observable outcome possible. Observable achievement should not be treated as evidence of independent functional capacity without identifying the support, adjustments, effort and recovery required to achieve it.
• v. Actively elicit hidden or normalised need. The methodology should be capable of actively eliciting needs that may be hidden, compensated for or normalised over time.

Some Autistic people under report functional difficulty because of masking, literal interpretation of questions, acquiescence, difficulty identifying internal states, learned adaptation, or because substantial family or other support has become their normal baseline. The assessment should not depend upon a person’s ability to independently identify and articulate every area of disability related need. This must be done while preserving the person’s voice, autonomy and supported decision making, rather than substituting another person’s account for their own. Otherwise, the characteristics of a person’s disability can themselves become a barrier to demonstrating disability related functional impact.

• vi. Use multiple forms of evidence. Accept diagnostic and specialist reports, functional assessments, observations, school and employment information, family, carer and kin accounts, participant narratives, support records and other longitudinal evidence. A standardised component may form part of the evidence, but it must never determine eligibility on its own, and decision makers must be able to consider broader allied health reports and other evidence where these provide a more complete picture of functioning across settings and over time.

A standardised component may form part of the evidence, with the weight placed upon it reflecting its demonstrated validity for the relevant population and circumstances. Where limitations or material uncertainty arise, decision-makers should consider broader clinical, functional, observational and longitudinal evidence and, where necessary, appropriately qualified professional judgement.

No applicant should be excluded solely because they cannot afford a particular private report.

• vii. Consider co-occurring conditions and intersectionality cumulatively. Autism frequently co-occurs with intellectual disability, psychosocial disability, communication disability, physical disability and other health conditions, and functional impacts cannot always be meaningfully understood in isolation. The framework must allow decision makers to consider the combined impact of multiple conditions and of intersecting disadvantage on functioning and participation, rather than requiring impacts to be attributed to a single impairment. Diagnostic overshadowing, in which distress or a change in functioning is attributed to autism when its cause is pain, ill health, medication, communication barriers or the environment, is a related risk that assessor training must address.
• viii. Build specialist workforce capability. Assessors and decision makers should hold relevant professional competence and disability specific training. For autism this includes communication differences, masking, sensory processing, interoception, executive functioning, demand related distress, Autistic burnout, co-occurring conditions and variable presentation across age, gender and culture. Capability should be backed by competency requirements, supervision, decision calibration, file audit and independent quality review, not simply completion of autism training, with monitoring of whether particular cohorts are being disproportionately refused or overturned on review.
• ix. Match reliance to validation. Assessment methods, thresholds and decision rules should be relied upon only to the extent that their validity has been demonstrated for the populations and circumstances in which they are used. Validation should specifically examine false negatives, masking, context dependent functioning, assessment modality, the effect of existing formal and informal supports and whether assessor capability

materially changes outcomes, across Autistic people with different communication, cognitive and support profiles, very high support and dependency needs, and assessor effects. The question is not only whether the tool is reliable, but who the tool gets wrong. Where validity is uncertain or an elevated false-negative risk is identified, this should trigger additional evidence or appropriately qualified professional judgement rather than automatically preventing implementation for populations for whom validity has been demonstrated. Validation and outcome data should inform whether application is expanded, modified or restricted. The Alliance’s prior evidence work on functional assessment, including the limited published base for the I-CAN, the limited autism specific evidence for the WHODAS 2.0 and the lessons of the 2021 independent assessments, is summarised at Attachment A, section 10, and the autism specific safeguards that follow from it are set out at Attachment C.

• x. Keep diagnostic evidence in its proper place. Diagnosis should neither automatically determine nor prejudice NDIS access, but relevant diagnostic and disability specific evidence about the nature, permanence and expected course of an impairment remains part of a multi source assessment and informs how functional evidence should be assessed. The reform should not move from over reliance on diagnosis to under reliance on it.
• xi. Separate assessment from rationing. Eligibility decisions should apply the law and the evidence, not assumptions about Scheme cost, service availability or what another system should provide. Scheme boundaries should be addressed transparently, after functional impact and permanence have been properly understood, and never by understating either. Access reforms should be driven by equity, evidence and support need; the consultation paper’s framing of the reforms within a commitment to limit Scheme growth makes it more, not less, important that the access process itself is insulated from that objective. Members’ community engagement indicates that Autistic people and families already perceive themselves as having been singled out in the public debate about Scheme sustainability, and that this perception is affecting disclosure and help seeking (Attachment A, section 9).33
• xii. Keep access separate from the level of support. Whether a person meets the disability requirements is a different question from what supports they need. Conflating the two, whether in tool design, thresholds or communication, risks people being refused access because their support needs are thought to be modest, or being allocated supports on the basis of an access assessment that was never designed for that purpose.
• xiii. Ensure procedural fairness. Provide the applicant with the information relied upon, an opportunity to correct errors or add context, accessible reasons for the decision and a clear, timely review pathway, and seek clarification before refusing access where evidence is incomplete. Human decision making, individualised assessment, supported decision making and effective review are the safeguards that the Robodebt Royal Commission showed are lost when high stakes decisions are automated or driven by population level rules.

Recommendation 6: Require autism and disability specific capability for everyone who assesses functional capacity or makes access decisions, backed by national competency requirements, supervision, decision calibration, file audit and independent quality assurance, not simply completion of autism training. Competency should cover communication differences, sensory processing, masking and camouflaging, executive functioning, interoception, shutdown, demand related distress, Autistic burnout, co-occurring conditions and the variability of presentation across age, gender and culture.

Recommendation 7: Functional capacity assessment methods should only be relied upon to the extent that their validity has been demonstrated for the population and circumstances in which they are used. This is consistent with the Alliance’s recommendation to the Senate inquiry and the lessons of the abandoned 2021 independent assessments. Where validity has not been established, or there is an identified elevated risk of false negatives, the assessment should not be determinative and additional evidence or appropriately qualified professional judgement should be required. Validation should specifically examine false negatives, masking, context-dependent functioning, assessment modality, the effect of existing formal and informal supports, very high support and dependency needs, and whether assessor capability materially changes outcomes, across Autistic people with different communication, cognitive and support profiles. Validation and outcome data should inform whether application is expanded, modified or restricted. The question is not only whether the tool is reliable, but who the tool gets wrong (Attachment C).

d. How could the access decision-making process, including referral pathways, be designed to reduce administrative burden and make it easier and more equitable, and support a proportionate approach for people who are clearly eligible?

Administrative simplicity should be measured from the experience of the person, family and supporter, not only from the efficiency of government processing. Many Autistic people and families already face repeated requests to prove a lifelong disability despite no material change in their condition, and the reform will be judged by whether that burden falls. The Alliance recommends an end-to-end pathway with the following minimum requirements:

• i. Supported entry. A no wrong door contact point, accessible information about the criteria and the process, and timely access to an independent navigator or advocate. Information about rights, available supports, referral options and review pathways should be actively provided, not left for the person to find.
• ii. An evidence plan. A plain language explanation, with examples, of the criteria, the types of evidence and information that may demonstrate functional impact, the evidence already held, any gaps and who is responsible for obtaining further information, provided in accessible formats including visual supports and Easy Read where appropriate. Predictability is part of accessibility: people should understand what is happening, what evidence is required, how it will be considered, what happens next and what they can do if something goes wrong.
• iii. Reuse before recollecting. Use existing government held and applicant authorised information, including existing reports and assessments, wherever lawful and relevant. Do not require repeated reports that add no material evidence, and do not require people to navigate multiple complex assessment processes across the NDIS, Foundational Supports and mainstream systems to establish access to support. A streamlined system

should support access to whichever of the NDIS and Foundational Supports is appropriate through one process.

• iv. Proportionate assessment. Streamline cases where reliable evidence already establishes permanence and very significant functional impairment and use fuller, tailored assessment only where necessary to resolve a material evidentiary question. For people already in the NDIS, the Agency may hold years of evidence about permanence, functional impact and support need, and reassessment during the three year transition should not require people to prove again matters already well established unless there is a genuine evidentiary reason. For Autistic people with very high support and dependency needs, whose lifelong impairment and substantial dependency are well-established, reassessment should ordinarily be a confirmation, not a re-trial. At the same time, ‘clearly eligible’ must not become shorthand for visibly disabled: the streamlined pathway should turn on the strength of the evidence, not on whether a person’s disability is apparent in an assessment room.
• v. Accessible participation. Offer home, community, clinic, phone or video options; interpreters; communication supports including augmentative and alternative communication; flexible timing and breaks; supporter participation; and other reasonable adjustments, at no cost to the applicant. For many Autistic people an unfamiliar clinic on a single morning is itself a barrier to demonstrating everyday functioning.
• vi. A transparent decision. Reasons that show what evidence was accepted, how conflicting evidence was resolved and how each criterion was applied, so that the person and their supporters can understand the decision and exercise review rights.
• vii. Supported referral with a failure trigger and continuity. A referral is not an outcome. Where another service system is identified as appropriate, there should be an identified receiving pathway and clear responsibility for transition. Where referral is refused, materially delayed, inaccessible, unsuitable or fails to meet the identified need, this should trigger rapid escalation and resolution and, where necessary to prevent a material gap in essential disability support, time-limited continuity while responsibility is resolved.

The process should not transfer costs to applicants. Where the decision maker requires additional assessment beyond what the applicant already holds, it should be arranged and funded by government, so that evidence requirements do not become a capacity to pay test for NDIS access. Standard templates may improve consistency, but templates should guide relevant information rather than constrain professional judgement or exclude narrative evidence. Reducing burden is also a matter of first time decision quality: poor initial decisions generate internal reviews, tribunal matters and repeated evidence requests, and the investment in workforce capability and decision support recommended in this submission is the most direct way to reduce that downstream cost for both applicants and government.

Recommendation 8: Provide proportionate pathways that do not turn on the visibility of disability: streamlined confirmation where permanence and significant functional impairment are already clearly established on the evidence, recognising that ‘clearly eligible’ must not become shorthand for visibly disabled; for current participants, reliance on the years of evidence about permanence, functional impact and support need that the Agency already holds, so that no one is required to prove again matters already well established unless there is a genuine evidentiary reason; reuse of existing evidence before new evidence is requested; and government funded assessment where the decision maker requires more

than the applicant already holds, so that evidence requirements never become a capacity to pay test for NDIS access.

Recommendation 9: Design the process around the person’s experience: a no wrong door entry point. Ensure independent disability advocacy is available where vulnerability, communication barriers, decision complexity, an adverse decision or a failed transition creates a heightened risk that the person cannot effectively navigate or challenge the process. This includes a plain language evidence plan that tells people what is happening, what evidence is required, how it will be considered, what happens next and what they can do if something goes wrong; accessible participation through interpreters, augmentative and alternative communication, Easy Read, visual supports, flexible modes and timing and supporter involvement at no cost; reasons that connect the evidence to each legislative criterion; a timely review pathway; and continuity of existing supports while a review or referral is unresolved. This is particularly important where support coordination and connection arrangements are reduced, delayed or unavailable (current rollout schedule is June 2028). Independent disability advocates should be recognised not only as supports for individuals, but as an early warning mechanism for systemic problems, including inaccessible assessment processes, failed referrals, service gaps and recurring transition failures. Government should establish a mechanism through which these patterns can be escalated and inform continuous improvement.

e. What should be considered to ensure the access decision-making process, and referral pathways, work well for all disability cohorts?

A universal process should provide consistent rights and safeguards, with assessment methods tailored to different communication, cognitive, cultural and disability needs. Disability related functioning is not static, and processes should not rely on observations from a single assessment interaction, particularly for people with less visible, episodic or context dependent disability. The following considerations are particularly important (and should not be read as definitively mutually exclusive).

• i. Autistic people. Recognise masking and camouflaging, sensory and communication differences, executive functioning, interoception, fatigue, shutdown, burnout, demand related distress, variable skills across settings and the cumulative effect of co-occurring disability. Do not equate verbal ability, eye contact, academic achievement or short term task completion with low support need. Recognise the diversity of Autistic people and of their support needs, including Autistic people with very high support and dependency needs and complex communication needs, for whom the practical test is whether the process would still work for a person who is non speaking or minimally speaking, may use augmentative and alternative communication, cannot travel independently, requires a familiar supporter and cannot tolerate an unpredictable or highly sensory environment. Assessment should recognise these functional realities without treating the label itself as an automatic access criterion.
• ii. People with communication disability. Provide communication partners, augmentative and alternative communication, Auslan, Easy Read, visual supports, extra processing time and alternatives to interview based assessment. Behaviour should never be interpreted without first examining communication access and unmet need.
• iii. People with cognitive or intellectual disability. Use concrete and accessible processes, supported decision making consistent with the Australian Law Reform

Commission’s recommendations, and evidence from people who know the person well. Avoid tests that depend heavily on literacy, insight or abstract self report.

• iv. People with psychosocial, fluctuating, episodic, degenerative or invisible disability. Assess typical patterns, frequency, severity, unpredictability, recovery time and cumulative impact. Consider foreseeable deterioration and the risk of judging capacity only during a period of relative stability. For Autistic people, whose autism frequently co-occurs with psychosocial disability, this consideration and the first are inseparable.
• v. First Nations peoples. Co-design with communities; enable community controlled and place based pathways; recognise kinship and community context; ensure cultural safety and trauma informed practice; and avoid imposing forms of evidence that are inaccessible or culturally inappropriate. Design and implementation should be informed by meaningful engagement with First Nations Autistic people and their communities, who are a priority cohort under the National Autism Strategy.
• vi. Culturally and linguistically diverse communities. Provide qualified interpreters, translated and culturally adapted information, and assessors skilled in distinguishing disability related differences from language acquisition or cultural variation, recognising that disability may be understood differently across cultural contexts.
• vii. People in regional, rural and remote communities. Where an access decision relies on an alternative pathway in a regional, rural or remote community, its actual local availability should form part of the transition decision. Where the pathway is unavailable, materially delayed or inaccessible, this should trigger escalation and resolution rather than leave the person without an effective pathway to support. Do not require specialist evidence that is locally unavailable. Fund travel or outreach where necessary, use telepractice appropriately and accept evidence from local practitioners and supporters with relevant knowledge.
• viii. Children and young people. Use developmentally appropriate, family centred and education informed approaches. Ground decisions in current evidence of disability and functional impact, considered alongside developmental trajectory, existing supports and foreseeable support needs, without speculative assumptions that future intervention will remove disability; protect timely early intervention; and ensure that the boundary with Thriving Kids does not create delay, duplication or a cliff edge in support. Recognise that for many Autistic children functional impact becomes more apparent, not less, as demands increase.
• ix. Adults and older people. Recognise the effect of lifelong adaptation, late diagnosis, changing informal supports, ageing carers, employment history and transitions in housing or community participation, and the particular risk that a late diagnosed adult who has masked for decades will be assessed as capable on the basis of the very strategies that have exhausted them.
• x. Families, carers, kin and supporters. Treat their evidence as valid evidence of functioning across settings and over time, while preserving the person’s own voice, consent, privacy and supported decision making rights. The Alliance’s consultation with families and carers during development of the National Autism Strategy emphasised that support needs, and what meaningful participation looks like, differ substantially across the spectrum; the process should be designed to hear that.

Referral pathways should be clear, accessible and coordinated for people whose needs span disability, health, mental health, education and community services. The same triggered safeguard set out in section 4(d) should apply across cohorts: a referral is not an outcome, and no referral should occur without reception. The failure of that pathway should trigger escalation, resolution and, where necessary, continuity to prevent a material gap in essential disability support.

Recommendation 10: Co-design culturally safe, community controlled and place based pathways with First Nations people and culturally and linguistically diverse communities; not requiring specialist evidence that is unavailable locally in regional, rural and remote areas; treating family, carer, kin and supporter evidence as valid evidence while preserving the person’s voice, consent and supported decision making rights; and never equating verbal ability, eye contact, academic achievement or short term task completion with low support need.

Recommendation 11: Where another system is identified as appropriate, there should be an identified receiving pathway capable of meeting the person’s relevant needs, with clear responsibility for referral, reception and resolution. A referral is not an outcome: an effective transition requires referral, acceptance and timely access to appropriate support. Where a referral is refused, materially delayed, inaccessible, unsuitable, fails to provide the identified support, or responsibility between systems is disputed, this should trigger rapid escalation and resolution and, where necessary to prevent a material gap in essential disability support, time-limited continuity while responsibility is resolved. No person should be required to resolve a disagreement between systems themselves. Governments should publish sufficient information about the availability, eligibility, capacity and wait times of the pathways being relied upon so that readiness and emerging gaps can be assessed. Accountability for the outcome should rest with an identified government decision-maker rather than be diffused across systems, with it possible to answer at each interface who is responsible, for what, from when, until when, what happens when the pathway fails and who owns escalation and continuity. Governments should monitor failed referrals, wait times and unresolved transitions so that recurring failures trigger system-level correction rather than repeated individual exceptions.

f. What factors and evidence should be considered when determining whether an impairment is permanent or likely permanent, including whether appropriate treatment is likely to improve the functional impact of an impairment?

Permanence should be determined through a balanced consideration of the nature and expected course of the impairment, treatment evidence, functional history and ongoing support need. It should not be reduced to whether any intervention might improve functioning, and the decision maker must keep four things distinct: the permanence of the impairment, functional capacity at a particular point in time, functional trajectory, and the support required to participate. Relevant diagnostic and disability specific evidence remains part of this assessment, because diagnosis is relevant evidence about the nature, permanence and expected course of an impairment. Relevant evidence includes:

• diagnostic and specialist evidence describing the impairment, its onset, expected course and co-occurring conditions;

• functional evidence across domains, settings and time, including support intensity and environmental adjustments;
• treatment and intervention history, including response, adverse effects, barriers, contraindications and whether further treatment is likely to remedy the impairment, as distinct from improving participation;
• longitudinal accounts and records from the person, family, carers, kin, educators, employers and practitioners;
• evidence of episodic, degenerative or fluctuating patterns, including frequency, duration, severity and recovery; and
• evidence of what occurs when support is unavailable, reduced or changed.

For autism, the guidance that accompanies the new process should make the following explicit. Autism is a lifelong neurodevelopmental disability; improvement in skills, participation or independence does not mean the underlying impairment is no longer permanent. Capacity building is not cure, and effective intervention is not evidence that support is no longer required. The existence of a theoretical intervention does not mean a person must undertake it before permanence can be established; treatment relied upon in an access decision should be evidence based, clinically appropriate, available, accessible, affordable, safe and reasonable for that particular person. People should not be required to undergo invasive, risky, repetitive or clinically inappropriate treatment, or to exhaust every intervention, before permanence is accepted. Successful adaptation or support is not evidence that the impairment has ceased; it may be evidence that current support is effective and necessary. The test should distinguish ordinary management and capacity building from treatment that is likely to remedy the impairment. Where an adverse decision rests on the proposition that treatment is likely to improve functional impact, the decision maker should be required to identify the treatment, the evidence of likely benefit, its accessibility, risks and likely timeframe, and why it is reasonable for that individual. Any consideration of treatment should be reasonable, evidence based and proportionate, and should not create unrealistic expectations of the outcomes that may be achieved.

Children require particular care because of the risk of circular reasoning: a child improves with intervention, permanence or significance is questioned, NDIS access is declined, and the child is directed to Thriving Kids or Foundational Supports. Improvement in a child’s skills, communication or independence is the purpose of early intervention, and the Australian evidence shows that its benefits are largest when it is timely. It is not evidence that the child’s autism is temporary or that support will not be needed as demands grow. Nor should a snapshot of current functioning be the whole picture: decisions should be grounded in current evidence of disability and functional impact, considered alongside developmental trajectory, existing supports and foreseeable support needs, without speculative assumptions that future intervention will remove disability. The recognition of autism as a lifelong condition in the NDIA’s current access lists should carry into the new guidance rather than be lost with the lists themselves.

Recommendation 4: Retain recognition of autism as a lifelong neurodevelopmental disability and keep four things distinct in legislation, rules and guidance: the permanence of the impairment, functional capacity at a particular point in time, functional trajectory, and the support required to participate. Clarify that capacity building is not cure; that improvement in skills, participation or independence through treatment, therapy, support, education or adaptation does not make an impairment temporary or show that support is no longer

required; that no one is required to undertake or exhaust treatment before permanence can be established; that treatment may be relied upon only where it is evidence based, clinically appropriate, available, accessible, affordable, safe and reasonable for that particular person; and that any adverse decision relying on treatment must identify the treatment, the evidence of likely benefit, its accessibility, risks and likely timeframe, and why it is reasonable for that individual.

Recommendation 5: Ground access decisions for children in current evidence of disability and functional impact, considered alongside developmental trajectory, existing supports and foreseeable support needs, without speculative assumptions that future intervention will remove disability. Early intervention is intended to improve functioning, and improvement must not itself be used as a reason to question permanence of a lifelong impairment or the significance of functional impact, decline access and redirect a child to Thriving Kids or other Foundational Supports. Where Thriving Kids or another Foundational Support is identified as the appropriate pathway, the transition should be supported and the receiving pathway identified. Where that pathway is unavailable, inaccessible, unsuitable, materially delayed or unable to meet the child’s identified needs, this should trigger escalation and resolution and, where necessary, protection against a material support gap.

Recommendation 12: Involve Autistic people, families and autism representative organisations in the design, testing, implementation and evaluation of the new process, including advice to the Technical Advisory Group and lived experience governance; publish disaggregated access and review outcomes, including review and overturn rates, by disability cohort, age, gender, geography and cultural background, with independent evaluation; set predefined indicators for identifying systemic inequity and equity and quality thresholds with authority to pause, modify or withdraw tools and decision rules that produce systematic inequity; and undertake and publish a cumulative impact assessment of the access reforms alongside the other NDIS changes commencing between 2026 and 2028.

5. Conclusion

The move from diagnosis lists to an assessment of functional capacity could make NDIS access clearer, fairer and more consistent for Autistic people, or it could reproduce inequity in a different form. The outcome will depend on design choices that remain open: whether functional capacity is assessed as it is lived or in a standardised room; whether the assessment methodology is independently validated for its intended purpose and for the populations and circumstances in which it will be used, and whether the degree of reliance placed on it reflects that evidence of validity; whether permanence is assessed as a question about the impairment or about whether treatment or supports are helping; whether the access process can identify and respond when assessment, decision-making or transition is at material risk of failure particularly where a participant is asked to leave the Scheme before the alternative(s) are demonstrably ready.

The Alliance’s recommendations are directed at those choices. They seek a process that makes the invisible visible and a structured professional judgement model built on multiple sources of evidence, autism capable assessors, published and validated methods, proportionate pathways, procedural fairness, a no gap transition guarantee, and transparent, disaggregated reporting with the authority to correct what the data reveal, tested at every point for safe transition arrangements.

The Alliance is not proposing more process for every person. It proposes a proportionate system that streamlines decisions where reliable evidence already establishes the relevant facts and triggers additional protection where it does not. For Autistic people with very high and established support dependency, this should mean confirmation rather than repeatedly proving lifelong disability. For people at risk of being missed by a standardised assessment, it should mean additional evidence or appropriately qualified professional judgement. Where another system is appropriately relied upon, transition should proceed; where that pathway fails, rapid resolution and, where necessary to prevent a material gap in essential disability support, targeted continuity should be triggered.

This approach also directs assessment and administrative resources to where they are most needed: avoiding unnecessary reassessment of established matters while concentrating additional scrutiny where evidence is uncertain, assessment limitations create material risk or transition is failing.

Access reform should therefore be assessed as an end-to-end system, not simply as an assessment methodology. Government should know where failure is foreseeable, detect when it occurs, protect the person from a material support gap and correct recurring system failures. The objective is not more safeguards everywhere. It is the right safeguard at the point of risk.

The Alliance would welcome the opportunity to discuss any aspect of this submission, to contribute to the work of the Technical Advisory Group on NDIS Functional Capacity, and to participate in the further consultation the Department has foreshadowed on thresholds, transition arrangements and implementation. Please do not hesitate to contact us with any queries or requests for further information.

Contact: chair@australianautismalliance.org.au
Jenny Karavolos, Independent Co-Chair, Australian Autism Alliance | jenkaravolos@outlook.com
Tade A, Senior Policy Officer, Australian Autism Alliance | temiediting@gmail.com

Attachment A: What Autistic Australians have told us and what the evidence shows about functional capacity, masking, burnout and permanence

The Alliance’s recommendations are informed by a substantial body of Australian autism community evidence gathered over several years, including its national consultation of more than 3,800 Autistic people, families and carers undertaken by Australian Catholic University for the Senate Select Committee on Autism, subsequent lived experience work undertaken for the NDIS Review, and consultation with families and carers during development of the National Autism Strategy, together with the Australian and international research relied on in the Alliance’s submissions to the Senate inquiry into the NDIS Amendment (Securing the NDIS for Future Generations) Bill 2026.32 This attachment summarises the evidence most relevant to the design of an access decision-making process.

1. Scale and relevance

Autistic people are the largest primary disability cohort within the NDIS. As at 31 March 2026, 338,099 active NDIS participants had autism recorded as their primary disability, representing approximately 44 per cent of active participants.1 Under the NDIA’s current operational guideline, autism assessed at DSM-5 Level 2 or 3 is a List A condition, likely to meet the disability requirements on diagnostic evidence, while other autism diagnoses sit on List B and require further evidence of functional impact.2 The consultation paper proposes that the lists be replaced from 1 January 2028 by a standardised, evidence based assessment of functional capacity, with current participants reassessed over a three year transition.3 Primary disability data do not capture the full intersection between autism and other disability cohorts. The scale and diversity of the Autistic population warrant an explicit autism lens in the design, validation and monitoring of the new process.

2. Functioning is context dependent

The World Health Organization International Classification of Functioning, Disability and Health, the international standard for describing functioning and disability, treats environmental factors as part of the classification itself: functioning and disability occur in a context, and change when environments change, independently of impairment.4 The National Autism Strategy recognises that environments, structures and practices can themselves create barriers for Autistic people and that support needs can vary according to circumstances and environments.5 The international consensus process that developed ICF core sets for autism concluded that no standardised, internationally accepted tools exist to assess autism related functioning.7

The Alliance’s national consultation shows how much of Autistic Australians’ everyday functioning is shaped by the environment around them. The Senate Select Committee on Autism reported that 47.5 per cent of Autistic adults sometimes felt unable to leave home because of fear of negative reactions; 41.4 per cent reported experiencing discrimination or stigma; 67.5 per cent of Autistic adults and 73.4 per cent of parents and carers reported feeling socially isolated; and 61.2 per cent of Autistic adults had lost friends because of the way others responded to their autism.6 A person who cannot leave home on many days is not well described by what they can do in a quiet room with a patient assessor.

3. Masking and camouflaging

Camouflaging, the conscious or unconscious masking of Autistic characteristics, is effortful and costly, with consequences including exhaustion and threats to self perception.8 A systematic review of 29 studies found that camouflaging is measurable and that higher camouflaging is associated with worse mental health.9 Autistic women report higher camouflaging than Autistic men, which makes under assessment a gendered fairness issue.10 Camouflaging and the number of unmet support needs each uniquely predicted suicidality in a study of Autistic adults.11 For access design, the implication is direct: an observation based assessment in an unfamiliar, structured setting will systematically see the mask rather than the person, and the people it misjudges most are those already at greatest risk.

4. Autistic burnout and fluctuating capacity

Autistic burnout is now well characterised in the research literature, including in Australian work. It arises from cumulative life stressors and the demands of masking and involves chronic exhaustion, loss of skills and reduced tolerance to stimulus, without any new impairment.1213 Burnout episodes may be short or long, recur, and are frequently misread as psychiatric illness.14 Capacity measured on one day is therefore a moving target, and an assessment that treats a period of relative stability, or a period of burnout, as representative will be wrong in opposite directions for different people. The design response is longitudinal, multi source evidence and explicit recognition of fluctuating and delayed presentation.

5. Permanence, treatment and early intervention

Autism is a lifelong condition. The American Psychiatric Association (DSM-5-TR) describes autism as lifelong while noting that the need for services and supports varies among individuals;15 the World Health Organization notes that some Autistic people live independently while others have severe disabilities and require lifelong care and support;16 and the National Institute for Health and Care Excellence maintains a standing clinical guideline for diagnosing and managing autism in adults, reflecting persistence across the lifespan.17 Support and treatment are concurrent with autism, not a route out of it: co-occurring mental health conditions are far more common in autism, and mental health assessment should be integrated into clinical practice.22

The evidence on early intervention supports timely access rather than a treatment first model. An Australian randomised controlled trial found that pre-emptive intervention in infancy reduced clinical autism diagnoses at age three by two thirds, with the authors noting that the first two years of life are when the brain is rapidly developing;18 long term follow up of the PACT randomised controlled trial provided the first evidence of long term symptom reduction roughly six years after early intervention;19 and a meta-analysis of 130 samples found significant positive effects for developmental and naturalistic developmental behavioural interventions, holding under analysis restricted to randomised trials, while noting that effect sizes attenuate under strict bias controls.20 The NDIA commissioned evidence synthesis found that effective intervention during childhood plays an important role in promoting learning and participation in everyday life.21 The corollary is that missed developmental windows are not recovered, and that improvement achieved through intervention is the intended outcome of support, not evidence that the impairment was temporary.

6. Diversity, co-occurring conditions and heterogeneity

Autistic Australians are not a homogeneous population. The diagnostic label masks a wide degree of heterogeneity between and within individuals at multiple levels of analysis,23 and the Lancet Commission on autism concluded that autism is heterogeneous and requires personalised, evidence based assessments and interventions.24 Autistic people may also have intellectual disability, psychosocial disability, communication disability, physical disability and other co-occurring conditions, and co-occurring mental health conditions are the norm rather than the exception.22 The National Autism Strategy recognises intersecting identities and circumstances that create additional and compounding barriers, with priority cohorts including First Nations Autistic people, culturally and linguistically diverse and culturally and racially marginalised communities, women, girls and gender diverse people, LGBTQIA+ people, regional and remote communities, and people with very high support needs.5 An access process built on population averages, or on a tool validated in a general disability sample, will produce very different experiences for particular groups, which is why validation across cohorts and disaggregated reporting are central to the Alliance’s recommendations.

7. Lessons from standardised assessment

Australia has direct experience of a standardised functional assessment model for NDIS access. Independent assessments were abandoned in July 2021, and the Joint Standing Committee on the NDIS found that standardised assessment tools may not be appropriate for certain cohorts of people with disability, and that the tools lacked relevance, sensitivity and specificity for the purpose of informing funding decisions or plans.25 The published validation record of candidate tools in autism remains narrow: the WHODAS 2.0, for example, has one published autism validation, in 109 Australian Autistic adults without intellectual disability using self report only, and its psychometrics have not been explored extensively in autism.27 Internationally, the United Kingdom’s Work Capability Assessment reassessed more than one million people using a functional checklist and has been the subject of peer reviewed research into associations with suicides, mental ill health and antidepressant prescribing.26 These experiences do not argue against functional assessment. They argue for publishing the methodology, validating it for the populations and circumstances in which it will be used, cohorts it will be applied to, testing it for false negatives, demonstrating workforce capability and matching the degree of reliance placed on an assessment method to the evidence supporting its validity for the relevant population.

8. Delay, decision quality and procedural fairness

The Commonwealth Ombudsman found that NDIA review delays pose a particular risk to people who may be at risk of losing services or experiencing deterioration, with a failure to triage urgent cases.28 The NDIS Review recommended new needs assessment processes that determine the level of need for each participant, an individual rather than class level decision logic.29 The Robodebt Royal Commission found that removing human oversight and reversing the onus of proof produced systemic, unlawful harm concentrated on vulnerable people.30 The Australian Law Reform Commission recommended that the will, preferences and rights of people requiring decision making support must direct decisions that affect their lives, and that supporters be recognised in the NDIS Act.31 Together these findings support the procedural safeguards recommended in this submission: individualised decisions, continuity of support while decisions and reviews are pending, supported decision making, accessible reasons and timely review.

9. Public discourse, stigma and disclosure

The climate in which the access reforms are being designed also matters to how they will work. In July and August 2026 Alliance member Amaze conducted a qualitative community survey of 30 respondents to inform the review of the Victorian Autism Plan, funded by the Victorian Department of Families, Fairness and Housing. Respondents included Autistic people, parents and carers of Autistic people and professionals, and reflected a diverse range of backgrounds including LGBTQIA+ people, First Nations people, people from culturally and linguistically diverse communities, people living in regional and remote areas and people with co-occurring conditions.33

Many respondents reported feeling that autism and Autistic people had been unfairly singled out in discussions about NDIS sustainability and reform, describing recent commentary as ‘dismissive’, ‘hurtful’, ‘dangerous’ and ‘utterly unfair’. Respondents consistently linked this rhetoric to increased stigma and negative community attitudes, and did not describe the impacts as merely emotional or symbolic: many reported that stigma and misunderstanding create real barriers to participation, support access and disclosure, with some feeling increasingly unsafe disclosing that they are Autistic. One respondent observed that ‘telling people I’m autistic feels the same as coming out in many ways, I can only do it if it’s safe’.33 The survey is small and qualitative, and the Alliance presents it as an indication of community sentiment rather than a population estimate. Its relevance to access design is direct: an access process depends on people being willing to describe their disability fully and to seek help, and a reform whose public framing deters disclosure will not produce accurate or fair decisions. This evidence underpins the submission’s recommendation that the guiding principles also govern how the reforms are communicated (section 4(b)).

10. The Alliance’s prior work on functional capacity assessment

This consultation should not be treated in isolation. The Alliance has already examined the proposed use of functional capacity assessment for NDIS access in three pieces of work in 2026, and that work provides an established evidence base for the positions in this submission.

The Alliance’s May 2026 Key Issues and Position Brief on the NDIS Amendment Bill examined the proposed functional capacity definition and the risks of assessing functioning without adequate regard to supports and environment.34 It identified Autistic functioning as contextual, fluctuating and support dependent, and highlighted sensory and environmental factors and support scaffolding. The risks identified included masking being interpreted as independence, family scaffolding being ignored, burnout being invisible and hidden disability being underestimated. The safeguards developed in that work included real world functioning assessment; consideration of cumulative fatigue and burnout, masking and camouflaging, support scaffolding, environmental and sensory impacts and fluctuating presentation; publication of the assessment methodology; inter rater reliability testing; cohort validation; access to raw assessment outputs; and review rights. The alternative approach identified was supported functioning, sustainable participation and real world functioning over time, rather than unsupported snapshot performance. The Brief also posed a question that remains directly relevant: how will assessors distinguish between sustainable functioning and functioning maintained through masking, exhaustion or informal family scaffolding?

The Alliance’s Supplementary Submission and Legislation to Evidence Matrix undertook a more detailed evidence review, testing the assumptions underlying the Bill against published research, authoritative reviews, Australian and international evidence and established evidence informed

practice.32 It identified significant evidence gaps, including the availability of an autism appropriate functional assessment methodology and the readiness of alternative support systems. For functional assessment specifically, the Alliance tested the assumption that a valid, reliable and fair methodology for Autistic people exists or would exist, and that functional capacity could be treated as a stable, context free attribute measurable through a point in time assessment. The evidence question expressly included inter rater reliability, autism specific validation, communication differences, masking, fluctuating presentation, context dependence and assessor competency. The evidence reviewed included the limited published base for the I-CAN and the absence, in the Alliance’s searches, of published autism specific validation, inter rater reliability or test retest research for that tool; Australian research involving 2,225 Autistic children showing that adaptive functioning may differ substantially from IQ; limited autism specific evidence for the WHODAS 2.0;27 and the international ICF work identifying the absence of an internationally accepted standardised tool for assessing autism related functioning.7 The Matrix also drew on the research on masking and camouflaging, gender differences in camouflaging, the WHO ICF recognition that functioning occurs in context, and the research on Autistic burnout summarised in sections 2 to 4 above, and on the 2021 Joint Standing Committee inquiry into independent assessments as Australian precedent.25 The resulting policy implications were publication of methodology, independent clinical validation, autism specific and cohort testing, inter rater reliability, workforce capability assessment, access to raw outputs and review rights before implementation, together with assessment of real world environments, masking, burnout and fluctuating presentation.

The Alliance’s June 2026 Senate Committee Brief placed functional assessment within the broader Systems That Work™ framework.35 It asked what evidence demonstrates system readiness and operational accountability; how harms would be identified and addressed in real time; what would happen if Foundational Supports were unavailable; what protections existed where families could not provide expected supports; how cumulative impacts had been assessed; what independent validation had been undertaken of the functional assessment framework; and what safeguards specifically addressed autism. The minimum preconditions it advanced included demonstrated readiness, accountability architecture, independent oversight, no gap protections, independent validation of the functional assessment framework, autism specific safeguards, assessment of family capacity and cumulative impacts, public reporting and independent evaluation. Its Systems That Work Readiness Index is a practical framework for determining whether replacement systems are sufficiently established before people are expected to rely upon them, and the accompanying Accountability Index addresses whether governments, agencies and service systems can be held accountable for outcomes, harms and implementation performance.

Taken together, this prior work provides an established evidence base for the current consultation: real world and sustainable functioning; masking and hidden disability; support scaffolding; context and fluctuation; communication and sensory differences; assessment validity; assessor competency; autism specific validation; system readiness; no gap protections; and monitoring and accountability. The current consultation is an opportunity to apply and further develop that work specifically to functional capacity assessment as an access gateway, and Attachment C does so.

Implications for the design of the access process

Taken together, the evidence suggests four questions should sit across the design, validation and monitoring of the new access process:

• Validity: does the assessment capture how the person functions in their real environments, over time, and with and without the supports currently in place, and is the degree of reliance placed on the assessment supported by evidence of its validity for the relevant population and circumstances?
• Permanence: is permanence assessed as a question about the impairment, with improvement through support understood as the purpose of support rather than as evidence that disability has resolved?
• Equity: are access and review outcomes examined separately for Autistic people, people with co-occurring conditions, people with very high support and dependency needs, people with complex communication needs, First Nations and culturally and linguistically diverse communities, and regional, rural and remote communities so that disparities can be identified and acted upon?
• Readiness and accountability: where another system is relied upon, is there an identified and appropriate receiving pathway; are failures detected and resolved; are baselines established and outcomes disaggregated; and can recurring failure trigger corrective action?

The access process should ultimately demonstrate not simply that decisions are consistent, but that they are sound: that people who meet the statutory requirements are identified, established matters are not unnecessarily reassessed, material support gaps are detected and addressed, and recurring failures lead to system correction.

Attachment B: Implementation and accountability tests for the NDIS access reforms

The Systems That Work™ Framework provides a practical way of testing whether the access reforms, as they are designed, validated and implemented, are producing the intended outcomes for Autistic people. The tests should be applied across the diversity of the Autistic community, including people with very high support and dependency needs and people with complex communication needs. The Alliance applies these tests in two ways: as a systems level test across the six domains of the access pathway, and through the five Systems That Work™ lenses. Access reform should be assessed as a system, not simply as an assessment methodology: a technically sound assessment cannot compensate for an unavailable alternative service, an inaccessible evidence pathway, inadequate assessor capability, a lack of procedural fairness or the absence of continuity safeguards.

The framework therefore tests three things at each stage: the baseline requirement that should apply to everyone; the trigger indicating heightened risk or failure; and the additional response required when that trigger occurs.

A systems level test across the access pathway

The five Systems That Work™ lenses then apply across the reform as a whole.

These lenses should be supported by a proportionate readiness test:

Is the particular assessment method or receiving pathway sufficiently valid and operational for the person or population to whom it is being applied; is there a defined trigger for identifying when that assumption fails; and is there an accountable response capable of protecting the person and correcting recurring failure?

The purpose is not to prevent or slow the access reforms. It is to ensure the reforms are evidence based, appropriately sequenced, monitored against outcomes and capable of correction. The ultimate test should therefore be:

Did the new access process identify the people who meet the NDIS access requirements, assess functioning as it is actually lived, avoid unnecessary reassessment where the evidence was already clear, detect when assessment or transition was failing, and act before that failure produced avoidable harm?

Attachment C: Autism specific assessment safeguards

As flagged in the response to question 3, the Alliance proposes the following autism specific safeguards for consideration in the design of the new access assessment. They build on the issues identified in this submission and on the Alliance’s earlier evidence and positions (Attachment A, section 10).

Any functional capacity assessment used for access should be designed and validated to:

• assess real world functioning across settings and over time, rather than relying on a single assessment encounter;
• recognise masking, compensation and learned adaptation;
• assess whether functioning is safe, reliable, repeatable and sustainable, including effort, fatigue, overload and recovery;
• distinguish observable achievement, supported functioning and independent capacity;
• identify formal, informal and environmental scaffolding, including prompting, supervision, co-regulation, routines and adjustments;
• assess what happens when existing support is absent or reduced;
• recognise executive functioning impacts, including initiation, sequencing, planning, transitions and task completion;
• recognise sensory processing, interoception, shutdown, burnout and demand related distress;
• distinguish communication from cognition and functional independence;
• recognise the assessment needs of Autistic people with very high support and dependency needs, including people with intellectual disability, complex communication needs and substantial reliance on others for daily living, personal care, health or safety;
• not require direct task performance or repeated functional testing where reliable longitudinal and supporter evidence already establishes very significant functional impact and dependency;
• provide augmentative and alternative communication, communication partners, additional processing time, written or asynchronous communication, breaks, multiple sessions and alternative assessment modes where required;
• actively elicit hidden or normalised functional impacts rather than relying solely on self identification of need;
• consider the cumulative impacts of co-occurring disability rather than assessing impairments in isolation;
• triangulate the person’s account with relevant longitudinal, diagnostic, clinical, functional and supporter evidence, with appropriate consent;
• treat reasonable adjustments as necessary to assessment validity, not merely to accessibility; and
• provide an opportunity to correct factual errors, missing context or an assessment conducted without necessary adjustments before an adverse access decision is made.

Any standardised assessment relied upon for NDIS access should be independently validated across the relevant Autistic populations and circumstances in which it will be used, including

people with different communication, cognitive and support profiles and people with very high support and dependency needs. Validation should specifically examine false negatives, masking, context-dependent functioning, assessment modality, the effect of existing formal and informal supports, and whether assessor capability materially changes outcomes.

The degree of reliance placed on the assessment should reflect that evidence. Where validity is uncertain or false-negative risk is elevated, additional evidence or appropriately qualified professional judgement should be triggered.

For Autistic people whose lifelong and very significant support dependency is already reliably established, the safeguard should ordinarily operate in the opposite direction: streamlined confirmation from existing evidence rather than additional functional testing. Proportionality requires both protection against under-identification and protection against unnecessary reassessment.

We should not only ask whether the tool is reliable. We should ask: who does the tool get wrong?

Reference list to Attachment A

1. National Disability Insurance Agency. (2026). NDIS Quarterly Report to disability ministers, Q3 2025-26 and associated appendices, reporting period ending 31 March 2026.
2. National Disability Insurance Agency. (2026). Applying to the NDIS: operational guideline, 26 August 2026, Appendices, List A and List B. (link)
3. Department of Health, Disability and Ageing. (2026). Consultation on NDIS access changes: public consultation paper, 31 August 2026. (link)
4. World Health Organization. International Classification of Functioning, Disability and Health (ICF). (link)
5. Australian Government Department of Social Services. (2025). National Autism Strategy 2025-2031. (link)
6. Senate Select Committee on Autism. (2022). Services, support and life outcomes for autistic Australians. Parliament of Australia, Canberra: Chapter 3, paragraphs 3.42 to 3.44; Chapter 6, paragraphs 6.105 to 6.107 and 6.146 to 6.150.
7. Bölte, S., Mahdi, S., de Vries, P. J., Granlund, M., Robison, J. E., Shulman, C., Swedo, S., Tonge, B., Wong, V., Zwaigenbaum, L., Segerer, W., & Selb, M. (2019). The Gestalt of functioning in autism spectrum disorder: Results of the international conference to develop final consensus International Classification of Functioning, Disability and Health core sets. Autism, 23(2), 449-467. (link)
8. Hull, L., Petrides, K. V., Allison, C., Smith, P., Baron-Cohen, S., Lai, M.-C., & Mandy, W. (2017). “Putting on my best normal”: Social camouflaging in adults with autism spectrum conditions. Journal of Autism and Developmental Disorders, 47(8), 2519-2534. (link)
9. Cook, J., Hull, L., Crane, L., & Mandy, W. (2021). Camouflaging in autism: A systematic review. Clinical Psychology Review, 89, 102080. (link)
10. Hull, L., Lai, M.-C., Baron-Cohen, S., Allison, C., Smith, P., Petrides, K. V., & Mandy, W. (2020). Gender differences in self-reported camouflaging in autistic and non-autistic adults. Autism, 24(2), 352-363. (link)
11. Cassidy, S., Bradley, L., Shaw, R., & Baron-Cohen, S. (2018). Risk markers for suicidality in autistic adults. Molecular Autism, 9, 42. (link)
12. Raymaker, D. M., Teo, A. R., Steckler, N. A., Lentz, B., Scharer, M., Delos Santos, A., Kapp, S. K., Hunter, M., Joyce, A., & Nicolaidis, C. (2020). “Having all of your internal resources exhausted beyond measure and being left with no clean-up crew”: Defining autistic burnout. Autism in Adulthood, 2(2), 132-143.
13. Higgins, J. M., Arnold, S. R. C., Weise, J., Pellicano, E., & Trollor, J. N. (2021). Defining autistic burnout through experts by lived experience: Grounded Delphi method investigating #AutisticBurnout. Autism, 25(8), 2356-2369. (link)
14. Arnold, S. R. C., Higgins, J. M., Weise, J., Desai, A., Pellicano, E., & Trollor, J. N. (2023). Confirming the nature of autistic burnout. Autism, 27(7), 1906-1918. (link)
15. American Psychiatric Association. (2022). What is autism spectrum disorder? (DSM-5-TR). (link)
16. World Health Organization. Autism fact sheet. (link)
17. National Institute for Health and Care Excellence. Autism spectrum disorder in adults: diagnosis and management. Clinical guideline CG142. (link)
18. Whitehouse, A. J. O., Varcin, K. J., Pillar, S., Billingham, W., Alvares, G. A., Barbaro, J., Bent, C. A., Blenkley, D., Boutrus, M., Chee, A., Chetcuti, L., Clark, A., Davidson, E., Dimov, S., Dissanayake, C., Doyle, J., Grant, M., Green, C. C., Harrap, M., … Hudry, K. (2021). Effect of preemptive intervention on developmental outcomes among infants showing early

signs of autism: A randomized clinical trial of outcomes to diagnosis. JAMA Pediatrics, 175(11), e213298. (link) 19. Pickles, A., Le Couteur, A., Leadbitter, K., Salomone, E., Cole-Fletcher, R., Tobin, H., Gammer, I., Lowry, J., Vamvakas, G., Byford, S., Aldred, C., Slonims, V., McConachie, H., Howlin, P., Parr, J. R., Charman, T., & Green, J. (2016). Parent-mediated social communication therapy for young children with autism (PACT): long-term follow-up of a randomised controlled trial. The Lancet, 388(10059), 2501-2509. (link) 20. Sandbank, M., Bottema-Beutel, K., Crowley, S., Cassidy, M., Dunham, K., Feldman, J. I., Crank, J., Albarran, S. A., Raj, S., Mahbub, P., & Woynaroski, T. G. (2020). Project AIM: Autism intervention meta-analysis for studies of young children. Psychological Bulletin, 146(1), 1-29. (link) 21. Autism CRC. Interventions for children on the autism spectrum: evidence synthesis commissioned by the National Disability Insurance Agency. (link) 22. Lai, M.-C., Kassee, C., Besney, R., Bonato, S., Hull, L., Mandy, W., Szatmari, P., & Ameis, S. H. (2019). Prevalence of co-occurring mental health diagnoses in the autism population: A systematic review and meta-analysis. The Lancet Psychiatry, 6(10), 819-829. (link) 23. Lombardo, M. V., Lai, M.-C., & Baron-Cohen, S. (2019). Big data approaches to decomposing heterogeneity across the autism spectrum. Molecular Psychiatry, 24(10), 1435-1450. (link) 24. Lord, C., Charman, T., Havdahl, A., Carbone, P., Anagnostou, E., Boyd, B., Carr, T., de Vries, P. J., Dissanayake, C., Divan, G., Freitag, C. M., Gotelli, M. M., Kasari, C., Knapp, M., Mundy, P., Plank, A., Scahill, L., Servili, C., Shattuck, P., … McCauley, J. B. (2022). The Lancet Commission on the future of care and clinical research in autism. The Lancet, 399(10321), 271-334. (link) 25. Joint Standing Committee on the National Disability Insurance Scheme. (2021). Independent Assessments. Parliament of Australia, Canberra. (link) 26. Barr, B., Taylor-Robinson, D., Stuckler, D., Loopstra, R., Reeves, A., & Whitehead, M. (2016). ‘First, do no harm’: are disability assessments associated with adverse trends in mental health? A longitudinal ecological study. Journal of Epidemiology and Community Health, 70(4), 339-345. (link) 27. Park, S. H., Song, Y. J. C., Demetriou, E. A., Pepper, K. L., Norton, A., Thomas, E. C., & Guastella, A. J. (2019). Disability, functioning, and quality of life among treatment-seeking young autistic adults and its relation to depression, anxiety, and stress. Autism, 23(7), 1675-1686. (link) 28. Commonwealth Ombudsman. (2018). Administration of reviews under the National Disability Insurance Scheme Act 2013: report on the National Disability Insurance Agency. (link) 29. Independent Review into the National Disability Insurance Scheme. (2023). Working together to deliver the NDIS: final report, recommendations and actions. (link) 30. Royal Commission into the Robodebt Scheme. (2023). Report. (link) 31. Australian Law Reform Commission. (2014). Equality, Capacity and Disability in Commonwealth Laws, ALRC Report 124. (link) 32. Australian Autism Alliance. (2026). Submission and Supplementary Submission to the Senate Community Affairs Legislation Committee inquiry into the National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Bill 2026. 33. Amaze. (2026). Survey to inform the Victorian Autism Plan review: findings relevant to the NDIS access consultation. Qualitative community survey of 30 respondents (Autistic people, parents and carers, and professionals), July to August 2026, funded by the Victorian

Department of Families, Fairness and Housing. Summary provided to the Alliance by Amaze, September 2026. 34. Australian Autism Alliance. (2026). NDIS Amendment Bill 2026: Key Issues and Position Brief, May 2026. 35. Australian Autism Alliance. (2026). Senate Committee Brief on the NDIS Amendment (Securing the NDIS for Future Generations) Bill 2026, including the Systems That Work Readiness Index and Accountability Index, June 2026. 36. Siegel, M., Lord, C. L., Tager-Flusberg, H., Ursitti, J., et al. (2026). Developing a consensus research definition for profound autism using a modified Delphi method. Molecular Autism, 17, 28.