NDIA Accountability Project

Submission 21 — South West Autism Network — New Framework Planning and NDIS Supports Consultation

‹ PrevPage 1 of 33 · Source p. 1Next ›

SWAN Submission to the Consultation on New Framework Planning Rules October 2026

DUE 14TH October 2026

Acknowledgements

SWAN acknowledges the traditional owners of the land on which this submission was produced, the Wardandi Noongar people. We acknowledge their deep spiritual connection to this land and extend our respects to community members and Elders past and present.

We also acknowledge people with disability and our families, for whom the National Disability Insurance Scheme (NDIS) was established.

About SWAN

South West Autism Network (SWAN) is a not-for-profit, charitable organisation that has been supporting autistic individuals and their families in the south-west region of Western Australia for the past 17 years. We are a Disabled Persons and Families Organisation (DPFO) with more than 2,000 members, and we provide free support to many more people with disability and their families. All staff, volunteers, and Board members either have a disability or are family members of someone with a disability.

Our primary role in the community is to provide information, peer support, advocacy, and connections to mainstream and disability services. We build the capacity of people with disabilities and their families to navigate disability and mainstream systems to meet their needs and participate in their local communities. We support people seeking diagnosis, post-diagnosis, and across their lifespan, and provide autistic-safe space group programs for autistic children, teens, and young adults through our peer support group programs. We also deliver Youth Mental Health First Aid training to the wider community.

As a regional not-for-profit Disabled Persons and Families Organisation (DPFO) providing information, peer support and advocacy, we are able to draw on 17 years’ experience supporting autistic individuals and their families, carers, support workers, allied health professionals, and the wider community. Our submission aims to include the voices of everyone who will be impacted by these proposed changes to NDIS assessments, planning and the NDIS Supports Lists.

SWAN Submission on New Framework Planning Rules October 2026 Page | 2

Contents

SWAN Submission on New Framework Planning Rules October 2026 Page | 3

Executive Summary

South West Autism Network (SWAN) supports a planning system that is fairer, clearer, accessible and genuinely person-centred. However, the information provided in this consultation does not demonstrate that New Framework Planning is ready to deliver those outcomes safely. Participation in this consultation does not signify endorsement of the proposed framework.

The Department of Health, Disability and Ageing has provided more information than in the previous consultation, but critical details remain unavailable. People with disability and our families cannot give fully informed feedback without the complete rules, assessment instruments, budget calculations and evidence showing how the proposed arrangements work together. Describing an assessment process as reliable and independently validated is not a substitute for publishing evidence that the actual modified tools and their intended funding uses are fit for purpose.

SWAN’s central concern is the absence of published, independently evaluated end-to-end pilot evidence demonstrating that the support needs assessment, report, budget calculation, plan approval and implementation processes produce accurate, adequate and usable plans. Testing individual components or the assessment conversation alone cannot establish whether the complete framework meets participants’ needs, especially given that this ‘testing’ was conducted with frontline NDIS staff not trained in the use of I-CAN v6 until very recently. Participants must be able to check their assessment reports and provide feedback at every stage of testing. There remains no end-to-end testing of the support needs assessment and new framework planning processes, and people with complex needs, children and young people have still not been included in the testing.

For autistic people and others with communication, cognitive, psychosocial or complex disabilities, an assessment conversation can significantly underestimate support needs. Masking, fluctuating capacity, sensory demands, executive-function disability, trauma and existing assistance can conceal the support needed in ordinary life. Being able to complete a task once, in a structured setting or with prompting, does not establish that a person can perform it safely and consistently without support. Furthermore, it is extremely common for autistic people, people with intellectual disability, acquired brain injury and psychosocial disability in particular to have a poor understanding of how their disability impacts their function and what supports they need, combined with significant difficulties communicating these impacts and needs on demand. Assessors must have relevant disability expertise, consider existing evidence, provide the accommodations necessary for meaningful participation, and have the skills to ask probing questions to accurately capture support needs.

The proposed budget calculation method must be transparent and responsive to individual circumstances. Default amounts, offsets and standard calculations must not override demonstrated need. SWAN opposes the proposed informal support offset. Family, friends and kin are not an unlimited substitute for funded assistance, and budgets must not create pressure for informal supporters to leave employment or provide care they cannot safely or financially sustain.

The promised flexibility must also be reflected in actual plans. Funding described as flexible provides little choice and control if it is largely restricted to specific supports, cannot supplement inadequate stated supports and is released through short funding periods. Participants need flexibility to respond to changing needs, equipment failures and loss of

SWAN Submission on New Framework Planning Rules October 2026 Page | 5

informal support without repeatedly applying for plan variations and/or reassessments. Self-management, direct employment and other self-directed arrangements must remain viable, with funding for their reasonable costs and employment obligations.

Regional equity requires more than a remoteness classification. Participants in isolated and underserviced towns face substantial travel costs and have no local providers, even where their location is not classified as remote. Budgets must recognise these circumstances, preserve provider choice and remain usable in practice. Unspent funding must not be treated as lower need when the real problem is unavailable services, restrictive plans or inadequate implementation support.

Changes to the support rules must preserve effective individual solutions. This includes suitable sensory supports, ordinary products that meet disability needs, in-home and daytime respite, meal assistance, disability-related exercise and health supports, and access to new technology. Replacement supports must not be removed until a workable alternative protects existing approvals and legitimate individual needs. Necessary support must continue where another system is unavailable or agencies disagree about responsibility.

SWAN is also concerned about the role of computer-assisted and automated decisions, and the subsequent use of sensitive assessment information. Participants need meaningful human accountability, access to the information and reasons behind decisions, and effective correction and appeal rights. Declining monitoring technology or exercising a right to complain, seek review or obtain advocacy must never count against a participant.

New Framework Planning must not proceed beyond voluntary, independently evaluated piloting until the complete framework is published, properly consulted on, validated for its intended uses and demonstrably safe. No participant should lose essential support or be made worse off because a new assessment method, calculation or administrative process fails to recognise their needs. SWAN’s recommendations seek enforceable safeguards that protect individualised support, genuine choice and control, regional equity and continuity of care.

Support Needs Assessments

Validation and fitness for purpose

The consultation paper states that the support needs assessment will be based on a modified version of I-CAN v6 and will assess 12 life domains. It will be supplemented by the PECQ, targeted assessments and assessed categories of NDIS supports. This is a high-stakes process because its outcomes will be used to determine what categories of support a participant is assessed as needing and, ultimately, their reasonable and necessary budget.

SWAN cannot support mandatory use of this approach without independent evidence that the complete, end-to-end assessment pathway is valid, reliable, accessible and fair for the populations to whom it will apply. It is not enough that a tool has been used elsewhere or has a history of development. The relevant question is whether the modified tool, delivered in this specific process and linked to a budget method, accurately identifies the real support needs of diverse NDIS participants. To-date, the I-CAN v6 has been validated for people with intellectual disability, but there is minimal evidence of validation for other disability types. Furthermore, as the NDIA are modifying the assessment tool, the validity of that tool is immediately compromised.

SWAN Submission on New Framework Planning Rules October 2026 Page | 6

Aggregate accuracy figures can hide poor performance for particular groups. The NSW Bureau of Crime Statistics and Research tested a machine learning model on a birth cohort of 259,160 people in Crime and Justice Bulletin No. 270 (Kim and Cheng, December 2025). The model was about 95 per cent accurate overall, but its precision sat between 32% and 55%, and its accuracy fell by up to 22% for Aboriginal young people. The authors concluded that ‘these results are a form of model bias … and they show that the aggregate performance of the model may be misleading as a justification for its use’. Published validation of the modified |-CAN v6, the PECQ and the budget model must therefore report performance separately for each disability cohort, First Nations people, children, and people in regional and remote areas, not only as a single overall figure.

This evidence must include autistic people across the spectrum of presentation, including people who mask or camouflage their disability; people with intellectual disability, psychosocial disability, acquired brain injury, sensory disability, degenerative conditions, complex communication needs, physical and neurological conditions and multiple disabilities; First Nations people; culturally and linguistically diverse communities; LGBTQIASB+ people; and people living in regional, rural and remote communities.

The support needs assessment must measure support needs across real settings and over time. It must not confuse observed capacity on a good day, in a structured environment, with a sustained capacity to complete tasks independently in everyday life. It must also capture the support needs that prevent crisis and build autonomy, including therapy, communication supports, sensory regulation, executive-function supports, support coordination, employment assistance and community participation.

The Department’s paper for allied health peak bodies proposes splitting therapy into ‘general’, ‘social’, and ‘physical and health related’ categories, each tied to the assessment domains it is meant to address. This proposal lacks the flexible approach recommended by the independent NDIS Review and promised by the Government. A person’s goal to parent safely, get back to work or take part in their community cannot be squeezed neatly into subdomains like mobility, interpersonal relationships, or learning and applying knowledge.

Assessor capability and independence

DHDA’s previous ‘What We Heard’ report records strong concern that assessors without allied health or equivalent disability expertise may miss clinically significant, functional and contextual information. SWAN agrees - the rules should set minimum capability standards, not leave this matter to general operational guidance.

The NDIS Review said that support needs assessments should be performed by ‘a skilled and qualified NDIA Needs Assessor with disability expertise who is a trained allied health practitioner, social worker, youth worker or similar’. The Department’s own summary of the February consultation records feedback asking ‘if non-allied health staff can use the tool for complex needs and how validation will happen’, and raising concern ‘that the tool will not cover daily support needs and miss capacity-building needs’. Both questions remain unanswered.

Assessors should have demonstrated competence in the participant’s disability type and ready access to specialist input. They must be thoroughly trained in autism, communication disability, psychosocial disability, intellectual disability, fluctuating and progressive disability, trauma-responsive practice, cultural safety, supported decision-making and accessible communication. They must understand that disability needs can be concealed,

SWAN Submission on New Framework Planning Rules October 2026 Page | 7

misunderstood or underestimated in short formal interactions, and be able to ask probing questions in order to properly identify and capture information about the participants’ support needs. It is extremely common for people with developmental and cognitive disabilities to inaccurately understand and express how their disability impacts their function, and what supports they need. For example, when asked what supports are needed to cook, a participant may answer that they can cook and don’t need support. When asked probing questions, it may then be identified that the participant has learned to cook toast, but is not able to cook a range of meals.

The rules should prohibit conflicts of interest and establish independent quality assurance, participant feedback processes and an accessible complaints pathway. Participants must be able to request a different assessor when there is a reasonable concern about safety, cultural appropriateness, gender, disability expertise, communication or previous harmful experience.

Individual evidence and participation

A participant should be able to provide reports and information from treating clinicians, allied health professionals, family, informal supporters, advocates, educators and service providers. The assessor must consider this material and record how it was treated. The NDIA should not place the entire burden on a participant to obtain costly new reports where existing evidence is available or where the NDIA needs specialist information to make a safe decision. Too many participants are unable to accurately and holistically explain their disability, how it impacts their function and what support needs they have; on demand and under the pressure of an assessment. Too many participants also have significant communication difficulties or no communication method — and where they also have no informal support able to provide a detailed explanation of their support needs, there is significant risk of a resulting inadequate budget unless pre-existing allied health reports are taken into consideration — particularly where support needs assessors lack allied health qualifications.

Participants need meaningful supported decision-making. They should be able to have a person of their choice present, use communication aids, take breaks, divide an assessment across multiple sessions, provide diary or video evidence, and respond in their preferred format. Whilst carers and supporters can assist a participant, but they must never be treated as a substitute for professional interpreters or as an unlimited source of support.

People have already told the Department what would help. As noted in the ‘What We Heard’ report, there are strong recommendations for ‘options for assessments to be conducted over multiple sessions, breaks to be accommodated, and for diary or video entries to be considered’. SWAN recommend putting these suggestions in the rules as examples of accommodations a participant can ask for and get, rather than leaving them to the discretion of the assessor on the day.

Draft report and replacement assessment

The draft support needs assessment report states that the participant does not need to do anything unless they think there are errors, and that the report will be used to work out their budget. This is insufficient. People should not have to detect and challenge errors after a high-stakes report has been written and is already moving towards budget determination.

The NDIS Review said the same, and so did SWAN in our submission to the Joint Standing Committee. The Justice and Equity Centre warns that if the support needs assessment report ‘is not provided to the participant before it is finalised, it could result in inappropriate

SWAN Submission on New Framework Planning Rules October 2026 Page | 8

budgets being set based on an inaccurate or incomplete needs assessment report*. Departmental material has since said participants will have the ‘opportunity to review the report and provide input and feedback before the CEO makes any planning decisions’. That commitment should be written into the rules, with a minimum number of days, so it is not lost when administrative processes change.

Participants must receive the draft report for checking before it is sent to the delegate, with at least 14 days to check it, correct errors, provide further information and add their own statement. If the draft report is sent to the delegate for review before being checked by the participant, it is highly likely that budget will be calculated at that point, risking any corrections by the participant being ignored. This is currently occurring on most first NDIS plans, where delegates are supposed to prepare a ‘draft’ budget prior to the planning meeting, then amend that budget after discussion with the participant. Instead, most new participants receive a two minute phone call from a delegate stating “this is how much funding you have, if you are not happy with it, get a report from an OT.”

The final report should show:

  • The participant’s strengths, goals and priorities.
  • Information and evidence considered, including participant-provided evidence.
  • Support needs for each relevant domain and sub-domain.
  • Raw scores and other assessment outcomes, explained in plain English.
  • How personal and environmental circumstances were considered.
  • Targeted-assessment findings.
  • The assessed support need categories, and any categories that were not assessed.
  • Any information that was not accepted and the reasons why.
  • The connection between the report and the later budget outcome.

A replacement assessment must be available promptly and without penalty where the assessment or report is inaccurate, incomplete, inaccessible, distressing, unsafe, culturally inappropriate, undertaken without necessary communication support, affected by an assessor conflict or lack of expertise, or fails to capture fluctuating, episodic, progressive or context-specific needs. A participant should not be required to exhaust complex review and appeal processes before a plainly flawed assessment can be corrected.

In the draft assessment report released with this consultation, the person’s ‘main informal support’ is listed as ‘Work or study full time’, which is not a support need at all. Under transfer and positioning the report scores ‘Frequency of support: Daily (3)’ and ‘Level of support: Managed (1)’, right beside a sentence saying the person ‘can independently transfer from one surface to another’. Walking and moving is scored ‘Never (0)’ and ‘Independent (0)’, while transport on the same page is scored ‘Daily (3)’ and ‘Extensive (4)’, and nothing explains how those varied results will be converted into a funding budget. Personal care is scored as independent with no support, even though the report says the person uses a visual aid in the bathroom to remember each step, and thus clearly requiring visual prompting support to perform this task.

When evidence is refused, the template says only ‘the information you gave us did not meet our requirements. We couldn’t use it to assess your support’. It does not say why, for that person. The template also tells assessors they ‘usually won’t need to complete this section where all recommendations have been funded’, so a person may be given reasons only when something has been refused, and even then in a set form of words. If the Department’s own example contains these errors, participants have little reason to be confident in the reports they will receive.

SWAN Submission on New Framework Planning Rules October 2026 Page | 9

Regional access

Regional and remote participants must have equivalent access to safe assessments. An online meeting cannot be treated as an equivalent default where a person prefers in-person assessment or where the person’s communication, sensory, cognitive, trauma or technology needs make video unsuitable.

Where a local in-person assessor is not available, the NDIA should fund reasonable travel and accommodation for the participant and their necessary support person, or send an appropriate assessor to the participant. Assessors working in regional and remote areas need local knowledge of service availability, distance, transport, culture and community context.

Budget Method and Participant Plans

Transparency before implementation

The consultation paper states that more detail about budget calculations will be provided later in 2026 after further testing. Whilst we welcome the commitment to release this information, it is already mid-October 2026, the commencement date for new framework planning is set to $1^{\text{st}}$ April 2027, and there has still been no end-to-end testing with participants.

The discussion paper says ‘further information and details about the budget method calculations will be provided later in 2026 after more testing is done’, then asks what information would help people understand how a budget is worked out. The Department’s July 2026 report already spelled out what happens when you consult this way: because explanatory documents for the assessment and budget rules were left out, ‘respondents were only able to provide feedback on how these rules are intended to operate, rather than how they will work in practice’. People asked the Department to publish ‘the methodology informing the budget rules, including how weightings and scores translate into reasonable and necessary budgets’. SWAN asks for that methodology again.

Most of the NDIA’s testing to-date was conducted by frontline staff not accredited in the use of I-CAN v6, and thus invalid. NDIA only ceased ‘testing’ with non-accredited staff in July-August 2026. Participants involved in the testing were not informed that the assessors were untrained frontline staff, and were not provided a copy of the resulting report for feedback. Even now that the NDIA is finally testing the ‘participant experience’ with accredited assessors, participants are still not being provided a copy of the resulting report to check accuracy. This lacks transparency, validity and accuracy. The lack of exposure drafts of the proposed new framework planning and budget calculation rules, only six months prior to the set date for commencement, and still with no end-to-end testing, is alarming. The potential repercussions of inaccurate support needs assessments cannot be understated - lives are literally at risk.

A budget method is not a technical detail. It is the mechanism that determines whether a person has enough funding to live safely, pursue goals, access therapy, maintain relationships, participate in community and avoid preventable crisis. Consultation is not meaningful unless people can see how assessment information translates into dollars.

The Department must publish the entire method, including the calculation for every support category; all benchmarks, averages, weightings, caps and assumptions; the role of any

SWAN Submission on New Framework Planning Rules October 2026 Page | 10

algorithm or automated process; the source and quality of the data used; all adjustments; how informal supports are treated; how regionality and thin markets are treated; and how the method responds to complexity, co-occurring disability, variable functioning and individual risk.

If the model determines all funding outcomes, the model itself should be published and open to review. The method should be independently reviewed by disability economists, statisticians, human-rights experts, people with disability and representative organisations. The review should assess disparate impacts, including whether particular disability cohorts, regional participants, women, First Nations people, culturally and linguistically diverse people or people with complex needs receive systematically lower outcomes.

Default amounts and deductions must not override individual need

SWAN is extremely concerned that default funding amounts could operate as caps in practice, even where an assessment identifies greater need. A default community-support amount must not replace an individual assessment of the assistance required for safe and meaningful participation.

The Department must also explain every proposed offset, including those involving aged care and vehicle modifications. An offset must not assume that another support is available, suitable or meeting the same need when this has not been established. The calculation must show what is being deducted, why, and how the Department has ensured that the same need is not deducted more than once.

Testing must examine the combined impact of default amounts, offsets, restrictions and funding periods. Each setting may appear reasonable in isolation while their combined effect leaves a participant without enough usable support. The rules must provide a clear remedy where the standard calculation does not meet the person’s assessed needs.

Informal supports

Informal supports must be recognised as voluntary, finite and separate from formal disability support. Family, friends, kin and community cannot be assumed to be available, capable or safe to provide personal care, supervision, transport, therapy support, crisis support or administrative work.

The rules must prohibit reducing an NDIS budget because of assumed informal support. Informal support should be considered only where the participant specifically wishes it to be considered, the person providing support agrees, the arrangement is safe and sustainable, and it does not replace supports the participant would otherwise reasonably need.

Attachment 4 of the paper proposes an ‘Informal supports offset: based on frequency and durability of informal support’. SWAN opposes this. The Department’s own report says people wanted informal supports treated as ‘voluntary and finite’, warned against assuming they are ‘the same as formal supports’, and asked the Department to ‘ensure informal supports do not determine the amount of funding a participant receives’. It also records that ‘some Aboriginal and Torres Strait Islander respondents were concerned informal supports may be over-relied on, with cultural obligations interpreted as unlimited caregiving capacity’. An offset calculated from the frequency and durability of family support is the outcome respondents asked the Department to avoid.

This is particularly important for carers who are ageing, unwell, working, parenting, geographically distant, experiencing family violence or burnout, or who have their own

SWAN Submission on New Framework Planning Rules October 2026 Page | 11

disability. It is also essential that kinship and cultural obligations are not misread as unlimited caring capacity.

Regional and remote loadings

Disability supports cost more and are harder to access in many regional, rural and remote communities. Participants face provider travel charges, limited workforce availability, limited or no registered providers, long waitlists, cancelled appointments and higher costs to obtain specialist assessment, therapy, equipment or home modifications.

The budget method must include mandatory and transparent regional and remote loadings, rather than discretionary adjustments. Loadings should account for direct costs as well as the practical cost of thin markets: travel, coordination, cancellations, provider scarcity, remote delivery, delayed repairs, need for flexibility and the inability to readily change providers. Importantly, loadings should also be applied to new framework plans of participants living in isolated towns that may not meet the remote criteria. For example, there are multiple towns in south west WA which have no registered or non-registered providers, and are located two hours or more away from the nearest provider. Participants in such locations must have sufficient funding to cover these additional provider travel costs.

The NDIA must never infer that lower spending means lower need without investigating whether the participant could access the support. Under-utilisation can reflect a lack of implementation support, market failure, waitlists, hospitalisation, inaccessible services, lack of transport, poor plan design or funding in a support category that is unavailable locally.

The Department’s budget fact sheet says only that ‘if a participant lives in a rural or remote location, there will be an adjustment to make sure they have the right amount of funding for their situation’. Its own report is clearer about the risk: ‘where services are limited, plans may appear adequate on paper but be unusable in practice’, and ‘lower utilisation rates should not lead to assumptions that support needs have lessened or reduced funding’. That second statement should be written into the rules.

Flexible funding, stated supports and restrictions

SWAN strongly supports genuine flexibility. The consultation paper says that flexible funding will be the larger part of the budget for most participants. The NDIS Review, the Government and the NDIA have consistently promised that new framework plans will primarily comprise of a flexible budget. This must be reflected in the rules and in the actual design of plans.

Stated supports should be restricted to circumstances where a specific item or service must be separately costed or safely managed, such as major home modifications, specialist disability accommodation or complex assistive technology. Participants must not be left with large amounts of stated or restricted funding that they cannot use because the required service is unavailable, unsuitable or unsafe.

The example New Framework plan illustrates the risks of prescriptive arrangements. The significant majority of the budget comprises stated or restricted supports, rather than the promised flexible budget, along with short funding periods and a registered provider requirement. While some safeguards may be appropriate for particular participants, restrictions must never be routine or imposed for administrative convenience. Any restrictions to the flexible budget must only be made at the request of participants or where there is significant evidence of misuse and the participant was provided substantive implementation support by the NDIA and/or NDIS partners

SWAN Submission on New Framework Planning Rules October 2026 Page | 12

The example new framework plan provided runs for three years, with a total budget of $338,400. In this example, $330,000 has been allocated as ‘flexible’ and $8,400 as ‘stated’. However, of the flexible funding, $240,000 is restricted, which leaves the participant with only $90,000 they have the choice and control to use flexibly. Only 26% of the example new framework budget provided is flexible. The reason given for restricting $240,000 is ‘this funding has been restricted because you have high in-home care needs. There would be a safety risk if you don’t receive these supports’. The plan gives no supporting evidence, no explanation of how the amount was worked out, no end date for the restriction and no way for the participant to have the restriction reviewed. The plan also tells the person ‘you can’t use any of your flexible funding to top up your stated supports’, and requires them to ‘choose an NDIS provider who is registered with the NDIS Quality and Safeguards Commission’. Given these further restrictions, the flexible budget is not, in fact, flexible.

Given that 74-80% of the budget in each example new framework plan provided by the NDIA to-date has comprised stated or restricted supports, we can only ask, where is the promised ‘majority flexible budget’?

New framework planning and support needs assessments have been sold to the disability community based on the promise of a primarily flexible budget, yet we have still not been shown even one example supporting this promise.

Any restriction or requirement must be grounded on individual, evidence-based reasons; be the least restrictive available option; time limited; reviewed regularly; explained clearly; and be open to review. The flexible budget must be truly flexible, including the ability to top up stated supports if they are inadequate, as is currently permitted in old framework plans. This is particularly critical for supports such as equipment repairs and maintenance. The participant must also be able to use another provider if the nominated or commissioned provider is unsafe, unavailable, lacks relevant expertise or is not meeting their needs.

Direct commissioning should be a last resort in thin markets. It must be co-designed and selected with local people with disability and communities, independently evaluated, and accompanied by clear safeguards against coercion. Participants must retain a meaningful right to change provider and access independent advocacy without fear that support will be withdrawn.

The new framework plan spending rules explanatory document allows the CEO to impose ‘a requirement that a participant utilise a certain provider in a remote First Nations community where the supports have been co-designed with that community through an alternative commissioning approach’. Before taking this approach, the CEO has to consider whether the requirement ‘would, due to local market conditions, substantially limit the participant’s access to the supports’. This test is problematic in two respects. A requirement that limits access, without substantially limiting it, remains permissible, and the CEO is only required to consider the question rather than be satisfied that the participant will in fact have choice and control. Furthermore, current commissioning arrangements do not include any local people with disability in the decision-making panel to determine which provider(s) is commissioned. This omission must be rectified in future commissioning arrangements, as we have already seen multiple providers with questionable reputations commissioned to service specific locations.

Commissioned providers should not be selected based on their ability to write a tender application. Service quality, history of complaints and poor performance must also be

SWAN Submission on New Framework Planning Rules October 2026 Page | 13

considered. Including local people with disability on the selection panel for commissioning arrangements helps address this issue.

The Department must also resolve the inconsistent treatment of plan management in the consultation material. Plan management is described as a stated support in Attachment 3, while Attachment 4 marks it ‘N/A’ and refers to possible legislative and commissioning changes. Participants need a clear explanation of what this means for their funding, choice of plan manager and ability to change arrangements.

The Department must also explain how support coordination, psychosocial recovery coaching and peer support will operate across the planning rules and related commissioning reforms. Necessary assistance must not fall between categories or disappear because responsibility is being moved to another arrangement. No change should interrupt an effective support relationship without a safe and workable alternative.

Protect self-management, direct employment and services for one

SWAN is concerned that the proposed rules do not clearly recognise the full costs of self-management, direct employment and running a service for one. Participants who employ their own workers have costs beyond the hours of support provided. These include recruitment, payroll, bookkeeping, superannuation, leave entitlements, insurance, employment advice, worker training, records, software and equipment needed to deliver support safely.

Larger providers include these costs in their prices. Participants who organise their own support must be able to fund the equivalent reasonable costs. Otherwise, the rules make self-directed support harder and push participants towards provider-led arrangements, regardless of whether those arrangements meet their needs or cost more.

Participants must also be able to set aside funding for employment obligations that are paid later, including leave entitlements. Money reserved for these obligations is not evidence of reduced support need. Funding periods and plan changes must not leave a participant unable to meet obligations already incurred in employing their support team. Suitable payroll and employment records must be accepted as evidence of expenditure, rather than requiring an invoice for each wage payment.

Self-direction is not limited to employing workers. It includes choosing independent workers, contractors, mainstream businesses and mixed support teams, subject to applicable legal requirements. It also includes peer mentoring, supported decision-making, help to recruit and manage workers, and disability-related assistance through microboards and circles of support. These arrangements help a person direct their own life, and must not be reduced to payroll services or replaced with provider-controlled arrangements.

Critically, participants must retain control of their individual funding. They must not be required to pool their budget with other participants or join a shared service as a condition of receiving necessary support.

Funding periods

Funding periods may help some people manage budgets, but rigid or short periods can and do cause harm. They can prevent participants from responding to cancellations, staff shortages, fluctuating and episodic needs, changing employment patterns, equipment delays, loss of informal support or a provider becoming unavailable.

SWAN Submission on New Framework Planning Rules October 2026 Page | 14

As consistently recommended by the Disability Representative and Carer Organisations (DRCOs), Independent Advisory Council (IAC), the Participant Reference Group (PRG) and the Better Planning Working Group, the default funding period should be 12 months. Shorter funding periods should occur only where the participant requests them or where there is clear evidence of overspend without reason. Any unspent amount should roll over to the next funding period until the end of the plan, and participants must be able to claim supports provided in a previous funding period from the subsequent funding period where needed. Disability support needs are rarely static in nature, and there are typically times when a participant needs to increase the use of funded support.

The Department already knows the risk. Its spending rules explanatory document says ‘if a participant requires support in irregular patterns due to the episodic nature of their impairment, shorter funding periods may not be appropriate as this will limit the amount of support they can access in a particular timeframe’, and that some people want longer periods so they have flexibility to meet their changing needs. The example plan is a three year plan where ‘funding will be allocated over 12 funding periods’. Whilst not explicitly explained in the documentation, the example again shows inappropriately short three month funding periods which, combined with 74% of the budget being stated or restricted supports, puts the participant at risk of not having their needs met.

Short funding periods must not be used as a substitute for appropriate plan implementation support, support coordination, navigation assistance or responsive communication from the NDIA.

Longer plans must retain their purchasing power

A longer plan only provides stability if its funding continues to be sufficient to purchase the support the participant needs. A three or five year budget must not be calculated using starting prices and then left unchanged while worker costs, provider prices and other necessary expenses increase.

The budget method must explain how funding will be adjusted over the life of a plan to preserve purchasing power. Participants must not have to reduce essential support hours or repeatedly request a plan reassessment simply because the cost of the same support has increased.

Longer plans must also include accessible, voluntary check-ins and a rapid pathway to increase funding where needs or service availability change. These check-ins should help identify unmet need, not become another compulsory assessment or a mechanism for reducing funding. The responsibility for recognising and correcting an inadequate plan must not sit entirely with a participant who has difficulty navigating the system. Furthermore, it is vital that the NDIA schedule check-ins where participants have not spent their funding or have significantly underspent their funding. This is a critical safeguard that rarely occurs in old framework plans. SWAN have supported dozens of participants and families who have not spent any of their NDIS funding for 6-18 months, and have not been contacted by any NDIS representative since the planning meeting. In each case, the participant (and their family) were in significant crisis by the time they were connected with SWAN.

Participant plan template

The draft new framework plan is clearer than many current NDIS documents in some respects, particularly in distinguishing flexible and stated funding. However, the template needs substantial improvement before it is used.

SWAN Submission on New Framework Planning Rules October 2026 Page | 15

Every plan should include a plain-English explanation of:

  • The participant’s total budget, plan length and funding periods.
  • Each flexible, restricted and stated amount.
  • Why each restriction, stated support or provider requirement was applied.
  • How the participant can request a variation or reassessment if their needs change.
  • How to ask for internal review and external review.
  • What evidence and calculations were used to set the budget.
  • How the participant can obtain support to understand and implement the plan.

New framework plans should also include a clear ‘How your budget was worked out’ section, supported by a detailed decision record. A person should not need specialist legal, financial or technical knowledge to understand whether their plan accurately reflects their assessed needs. Recorded webinars explaining new framework planning and support needs assessments available on demand would also support participants and families to understand and navigate these changes.

The example plan says ‘your goals are set by you and written in your own words’, then gives a goal that starts in first person and switches to third person halfway through. In our experience, old framework plan goals are typically (and quite obviously) written by either a NDIS partner or the NDIA delegate, and commonly do not reflect the participant (or their family’s) actual goals. Moving forward as new framework plans are developed and now that funding budgets are decoupled from plan goals, NDIA should prioritise including goals written in the wording chosen by the participant - unless the participant specifically requests that no goals are recorded on the plan. Of further note, assistive technology in the example plan comes with the note ‘there are no funding periods for this stated support’, which is sensible, but the plan never explains why some supports are treated that way and others are not.

Plan Variations, Reassessments and Review Rights

Plan variations must be responsive

SWAN welcomes an intention to use plan variations for short-term and limited changes, including urgent funding for crisis or safety reasons, repairs and replacements, and changes to employment. This needs to be strengthened in the rules.

The discussion paper says that ‘most plan updates will be made through a plan variation rather than a full plan reassessment’, and the fact sets out the difference: a variation changes part of a plan without making a new one, while a reassessment is a full review that produces a new plan. The Department’s ‘What We Heard’ report notes ‘strong support for ‘light touch’ variations and interim funding options that avoid full reassessment for minor or temporary changes, alongside the ability to review draft plans before approval to prevent avoidable variation requests’. SWAN supports this approach. However, it only helps people if the rules set time limits for decisions and a participant can ask for a variation themselves, instead of waiting for the NDIA to pick which path applies.

Plan variations must be available where a plan is unsafe, inadequate or unworkable in practice, even if the change is temporary. Examples include a carer being hospitalised (expected or unexpected); family breakdown; provider loss; a change in living arrangements; an urgent equipment repair; an increase in disability-related support needs after illness or injury; risk of neglect, abuse or homelessness; sudden loss of transport; or a participant needing additional help to start or retain work or study.

SWAN Submission on New Framework Planning Rules October 2026 Page | 16

The rules must provide rapid decision timeframes and interim funding where delay would create risk. A participant should not have to prove that a crisis will be permanent before receiving temporary support necessary to remain safe.

Importantly, an urgent plan variation is not enough if the participant has to wait for a decision before essential equipment can be repaired. A broken hoist, wheelchair or positioning device can prevent a person getting out of bed, moving safely or receiving necessary personal care or increase the risk of the participant asphyxiating. The current 21 day time frame for NDIA to respond to plan variation requests is too long to ensure safety. The rules must allow participants to use available flexible funding for urgent repairs, including where their stated equipment repair funding has been exhausted.

Plans must include realistic repair and maintenance funding, however, a stated support laid out by budget calculation rules cannot predict every breakdown or several equipment failures at once. The rules must provide a clear process for reimbursing necessary emergency repairs paid for by the participant or family, with evidence requirements proportionate to the cost and risk, and for individuals unable to pay out of pocket, the ability to use their flexible budget to pay for urgent equipment repairs. Furthermore, repairing or replacing essential equipment must not require an otherwise suitable plan to be reopened or expose unrelated supports to unnecessary reassessment.

Reassessments must not replace support

A full plan reassessment may be appropriate where support needs have changed substantially and ongoingly, but it should not be used as a barrier to a practical plan change. Nor should a participant who seeks an urgent variation be compelled to transition into New Framework Planning while in crisis.

People with lifelong disability should not be repeatedly reassessed when their overall support needs are stable. The process should be proportionate and should build on existing evidence, rather than requiring a person to repeatedly retell traumatic experiences or prove disability that has not changed.

Reviews and appeals

Participants must retain effective rights to seek internal review and Administrative Review Tribunal review. These rights are particularly important where a standardised assessment and budget method may make it difficult to understand the source of an inadequate outcome.

One decision cannot be challenged at all. The Transition to New Framework Plan explanatory document says ‘the notice to have a new framework plan is not a reviewable decision and once issued a participant will proceed through the new framework planning process’. The Notice of Impairments document says the opposite about impairment notices: ‘a decision to vary, or not to vary, a notice of impairment is a reviewable decision under section 99 of the Act’. SWAN can see no good reason for this difference. A decision that pushes someone into an untested or inadequately tested planning system, at a time the Agency picks, should be reviewable. At the very least a person should be able to ask for it to be postponed, on grounds set out in the rules.

The Department’s earlier ‘What We Heard’ report records strong calls for review processes that are simple, timely, accessible and independent; for a way to reconcile treating-

SWAN Submission on New Framework Planning Rules October 2026 Page | 17

professional evidence with assessment outputs; and for light-touch variations and interim funding where appropriate. SWAN strongly supports these calls.

The rules and operational guidelines should require:

  • Written, accessible reasons for every decision.
  • A clear pathway to correct errors without needing a full legal review.
  • Fast replacement assessments where assessment quality is in question, conducted by a different assessor chosen by the participant.
  • Direct contact with the participant during internal review unless they specifically opt out.
  • An internal review decision that addresses each issue raised.
  • Continued essential funding while review or appeal is underway where a plan cuts or interrupts existing necessary support.
  • Access to independent assessment and expert evidence through Administrative Review Tribunal appeal, where the NDIA assessment is disputed.
  • Public reporting on timeframes, outcomes, demographics and participant experience of reviews and appeals.

Exercising a right must not create a record that counts against a person. Requesting an internal review, lodging a complaint, engaging an advocate, making a freedom of information request, contacting a member of Parliament or changing provider must never be treated as a risk, fraud, compliance or behavioural indicator, and must never affect an assessment score, a budget, funding period timing, funding management type, or a decision to impose other restrictions. The Rules should explicitly state this, and the NDIA should publish confirmation that these variables are excluded from its models.

Changes to NDIS Support Rules

Clear, principles-based rules

SWAN supports clear information about what can and cannot be funded, but rules must not become so prescriptive that they remove sensible, individualised decision-making. The consultation paper says the revised rules will group supports differently and will more clearly identify certain items that the NDIS will not fund. It also proposes to remove replacement supports for both old and new framework planning, with some commonly requested items to be included as stated supports.

The Department’s paper for allied health peak bodies proposes three therapy categories, general, social, and physical and health related, each tied to assessment domains. A participant will be told which category they can use, ‘likely as part of the participant’s needs assessment report and/or plan’, but the report ‘will not necessarily identify the therapy type (for example speech pathology)’. The same paper says that under the new arrangements ‘most supports will no longer be described in a participant’s plan’, and that the categories are ‘still subject to policy refinement and consultation’. People are being asked to accept a set of categories that is not finished and that will not tell them which allied health professional(s) they are able to see.

SWAN is concerned that removing replacement supports without a broad, accessible and timely alternative pathway will create gaps for participants whose disability-related needs do not fit neatly in a fixed list. A highly prescriptive list cannot anticipate every safe, reasonable and necessary support. It can also lead to inconsistent frontline interpretation, administrative delays and people missing out because a support is not named or because the process is too difficult to navigate.

SWAN Submission on New Framework Planning Rules October 2026 Page | 18

The rules should be written in plain language and anchored in principles: the support must relate to disability support needs, be effective and safe, represent value for money, be appropriately funded by the NDIS rather than another system in practice, and support the participant’s autonomy, inclusion and goals. Lists should clarify obvious exclusions, not operate as a narrow substitute for individual consideration.

Saying that a support belongs to health, education, housing or foundational supports does not confirm that the participant can actually access it. People with disability do not fit neatly into boxes, and the siloing of support to different agencies risk gaps increasing that impact safety and wellbeing. Before necessary support is removed or transferred, the responsible agencies must identify a funded, accessible and suitable alternative for that person and arrange a working handover.

Where the alternative is unavailable, or agencies disagree about responsibility, necessary support must continue while the issue is resolved. There must be a named contact with authority to coordinate a response. Participants and families must not be left to pay privately, go without support or negotiate disputes between Government systems. This is particularly important in regional and remote communities, where a service that exists in a metropolitan area is commonly unavailable locally.

Respite must fit the participant’s needs

SWAN is concerned that describing short term accommodation as overnight time away from family or other unpaid supporters may exclude arrangements that are safer, more suitable and less expensive. Respite must also be available at home, during the day or overnight, with familiar workers where this is what the participant needs.

For many autistic people, people with intellectual disability, acquired brain injury, psychosocial disability and people with complex disability; leaving home means losing access to familiar routines, necessary equipment and trusted support. In regional and remote areas, suitable respite accommodation rarely exists. A person should not have to leave their home or accept unfamiliar workers for their family or other informal supporters to have a break.

The rules must also allow short term accommodation funding to be used flexibly as additional in-home support worker hours instead of out-of-home respite where this meets the participant’s needs. In addition to out-of-home respite being unsuitable and/or unsafe for some participants, since the changes to the NDIS operational guidelines for short term accommodation requiring participants to pay for food and activities, it is now completely unaffordable for people living in poverty and those with high non-NDIS disability costs (eg. medication and other health supports).

Whilst SWAN commends the more equitable approach banning claims for food and activities by providers as well as self directed STA, the new rules for STA in new framework planning must ensure that the flexibility to claim increased support worker hours or in-home respite using STA funding in old framework planning is also available in new framework planning. Without this change, more participants will experience breakdown in their relationship with their informal supports, more families will be forced to relinquish care sooner, and more participants will be at increased risk of violence, abuse and neglect.

SWAN Submission on New Framework Planning Rules October 2026 Page | 19

Replacement supports

Before replacement supports are removed, the Department must publish a detailed transition analysis showing which current items and services will no longer be available, who will be affected, what alternative pathway will apply, expected decision timeframes, and how urgent needs will be met.

Moving some smart phones, tablets, smartwatches and communication applications into stated support categories does not resolve access to ordinary household items or other individual solutions. A standard household product may meet a disability need safely and at lower cost than ongoing paid assistance. The rules must provide a workable pathway for these items, rather than requiring a disability-specific label or unnecessary modification.

Existing replacement support approvals must remain effective during transition, and pending applications must receive fair decisions rather than being discarded because the process changes. Participants who remain on old framework plans must also be able to access newly needed eligible supports. They must not be required to wait for New Framework Planning to obtain a support they need now. Furthermore, the rules must be sufficiently flexible to enable access to new and innovative technologies, apps and equipment.

Suitable low-cost, low-risk supports should be available through direct purchase within an adequate budget, with simple records of their disability-related purpose. Other, more complicated requests need timely, reviewable decisions without repeated reports or unnecessary reassessment of the whole plan.

A support must not have to replace a worker to demonstrate value. Some participants need both the equipment and the person assisting them. Value includes improved safety, communication, autonomy and participation, not just a reduction in paid support hours. The rules must also enable the trial of safe, low-cost solutions and individual community arrangements for people who do not fit a standard program.

The proposed distinction between assistive technology ‘Assets’ and ‘Other’ also needs clarification. The Department must explain which items fall into each category, whether their funding is flexible or stated, and what approval and evidence requirements apply.

A low-cost device should not become harder to fund simply because it needs occasional maintenance, a software update or a replacement part. These are ordinary features of many useful technologies. Funding arrangements must reflect cost, risk and individual need, rather than create another administrative barrier to effective support.

Participants should never lose access to a disability-related item or service simply because it is not included in a stated-support category or because the NDIA takes too long to make a decision. There must be a rapid, accessible pathway for unusual, emerging and individual needs, including technology, communication supports, sensory supports, low-cost assistive technology and safety-related items.

The specific proposed exclusions also require careful reconsideration. A category-based exclusion can have unintended consequences for people whose disability-related needs are unusual or highly individual. The decision should always consider the functional purpose, safety and disability-related need, not merely the retail label or commercial availability of an item.

SWAN Submission on New Framework Planning Rules October 2026 Page | 20

Electronic noise-cancelling headphones and sensory supports

SWAN strongly opposes a blanket exclusion of electronic noise-cancelling headphones. For an autistic person with significant sensory needs, these may be an essential disability support that makes it possible to leave home, use transport, attend appointments or participate in community life. Being an ordinary retail product does not remove their disability-related purpose. Many autistic people supported by SWAN rely on electronic noise-cancelling headphones because they are more effective and reducing noise than standard ear defenders, rather than using them to listen to music etc as is typical of non-disabled users. A standard household item is something used by the majority of the population, whereas noise-cancelling headphone use is still well within minority usage, as are smartwatches and other commonly used ‘mainstream’ assistive technology used by people with disability.

Any concerns about the use of electronic noise-cancelling headphones by young children must be assessed in that context. They do not justify excluding headphones for adults or other participants with different needs. The Department must publish the evidence and explain which ages, circumstances and uses it covers.

Funding decisions must consider the individual’s sensory needs, the purpose of the support, safety and value for money. Sensory difficulties are part of the DSM-V(TR) diagnostic criteria for autism, and a core feature of disability impact for many autistic people. Where electronic noise-cancelling headphones are an effective and appropriate disability support, the rules must allow them to be funded. Participants should not have to lose access to a relatively inexpensive support and then require more intensive assistance to manage the same environment, or experience increased isolation.

Exclusions must not contradict permitted supports

The proposed exclusions must distinguish ordinary household and personal expenses from disability-related modifications, access needs and support costs. Excluding fencing and gates must not also exclude a necessary access modification, gate automation or maintenance of an existing disability-related modification.

Similarly, excluding hair styling tools must not automatically exclude an adapted tool or an ordinary product that enables a person to manage disability-related personal care. The relevant question is what the item does for the participant, not whether non-disabled people also buy it.

The Department must also explain how the proposed exclusions of hydrogen fuel and electric vehicle charging interact with eligible transport and support worker travel. Equivalent disability-related travel costs should not be treated differently simply because a vehicle uses a different energy source.

Participants, plan managers and NDIA staff must not be left to make personal judgement calls on contradictions between an excluded item and a permitted disability support. The final rules must make these boundaries clear.

Preserve practical choices for daily living support

The final rules must expressly preserve disability-related meal preparation and delivery assistance, while distinguishing these support costs from ordinary food costs. A change in category wording must not remove a practical support that participants already rely on.

SWAN Submission on New Framework Planning Rules October 2026 Page | 21

Participants must be able to choose between suitable meal delivery, assistance with preparation and support worker assistance within their available eligible funding, without requesting a plan change each time. Some people need help to eat but do not want a worker in their kitchen. Others need different arrangements when their capacity fluctuates. Choice and control must include how an everyday disability need is met.

The broader in-home and community category must also preserve assistance with daily tasks outside the home, including grocery shopping. Combining categories should improve flexibility - not create new uncertainty about supports that were previously covered.

The Department must clarify whether the proposed therapy wording preserves disability-related personal training by suitably qualified trainers, as well as exercise physiology. Moving supports into a therapy category must not unintentionally restrict all assistance to allied health practitioners or remove an effective support because its provider has a different qualification.

Disability-related health support must have clinical oversight matched to the task, the risk and the participant’s needs. The rules must allow appropriately trained and capable support workers to perform suitable tasks, with access to the clinical guidance and supervision required. They must neither impose unnecessary continuous clinical supervision nor transfer tasks requiring a clinician onto workers without the necessary competence and support. Funding must cover training and the actual cost of high-intensity support.

The rules must also expressly preserve necessary disability-related communication and behaviour support when a participant accesses health or mental health services, including hospitals. Being in hospital does not make a person’s communication or behaviour support needs disappear, and it is unreasonable to expect health or mental health staff to be trained in providing this specialist support to patients with disability. Participants must not be left unable to communicate, understand their care or receive necessary assistance because two systems disagree about responsibility.

Accessible interpretation and implementation

DHDA must not repeat previous approaches where rules were difficult, confusing and overly complicated for participants, families, providers and NDIA staff to interpret. Inaccessible or unclear rules create underspending, unnecessary fear, mistaken denials and an increased risk of participants being penalised for honest errors.

Before implementation, the Department and NDIA should provide a complete suite of accessible materials: full plain-English guidance, Easy Read versions, Auslan videos, translated resources, screen-reader accessible documents, worked examples and a reliable escalation pathway for urgent clarification. Staff, partners, plan managers and providers must receive consistent training that reflects the legal rules and respects participant rights.

Participants need one official source of reliable funding advice, not conflicting (and often incorrect) answers from the NDIA, partners, plan managers and providers. The Department and NDIA must establish an accessible help desk that provides clear written answers, supported by practical examples and published decisions with identifying information removed.

Participants must be protected where they reasonably rely on official written advice. Likewise, participants must be protected where they have not been provided adequate

SWAN Submission on New Framework Planning Rules October 2026 Page | 22

implementation support to understand the changes and their obligations. They must not face debt recovery, loss of plan-management choice or other adverse action because the Agency later changes its interpretation or another staff member gives a different answer.

Evidence requirements must be proportionate to the cost and risk of the support. Existing evidence should be reused where the need is already established. Renaming a support or moving it to a new category must not, by itself, trigger a requirement for another assessment or report.

Transition, Consultation and Accountability

Transition must be safe and voluntary first

The Government now proposes transition to new framework planning from April 2027. The later start provides additional time, but time alone is not a safeguard. The transition must not begin on a compulsory basis until the tools, budget method, rules, workforce, participant supports, and review systems have been proven to work safely. This has not yet occurred, and given that there remains no end-to-end testing of the entire support needs assessment process, resulting report, new framework plan and budget calculation with participant feedback and evaluation on every step, modification of |-CAN v6 rendering the tool no longer validated, and no testing with people with complex needs, children or young people; SWAN is extremely concerned that preventable harm will occur to participants if commencement proceeds as scheduled.

The Department’s July 2026 report says ‘we heard clearly some aspects of the consultation process did not meet expectations’, that people wanted ‘more information about the support needs assessment and budget method rules, and more accessible formats’, and that new framework planning ‘will now start in April 2027’. The Government pushed the date back because the work was not done. The same test should be applied prior to April 2027. Before any group of participants is moved across, the Government should transparently confirm that the tools, rules, budget calculation, workforce and safeguards are fully (and validly) tested and effective for all disability cohorts, First Nations, culturally and linguistically diverse people, LGBTQIASB+, women, regional and remote, people with complex needs and children and young people.

There must be a full, end-to-end pilot of support needs assessments and new framework planning with independent evaluation. The first phase should be voluntary, independently monitored and designed so participants can compare outcomes without losing their existing plan protections. Participants who receive an inadequate New Framework Plan must be able to return to their previous funding arrangement or maintain prior funding while errors are corrected and review occurs.

No participant should be transitioned during acute crisis, hospitalisation, homelessness, bereavement, family breakdown, urgent safeguarding concerns, major life transition, loss of a key informal support, or while a material error in their Notice of Impairment or other NDIA record remains unresolved.

Minimum transition pathway

Before a person transitions, the NDIA should:

  1. Provide accessible information well in advance, in the participant’s preferred communication format.

SWAN Submission on New Framework Planning Rules October 2026 Page | 23

  1. Give the participant sufficient time to check and correct information the NDIA holds about disability, impairment, communication, culture, support needs and risks.
  2. Offer a preparation meeting with independent advocacy, navigator or support coordination assistance.
  3. Provide examples of the assessment report and participant plan.
  4. Explain the participant’s rights, the likely process, available supports, review pathways and options about assessment format and support people.
  5. Confirm that the participant is ready before issuing a transition notice.
  6. Conduct assessment in the participant’s preferred safe and accessible format. Participants must be able to access in-person assessment if that is their preferred method, regardless of their location.
  7. Give the participant their draft assessment report for correction before review by an NDIA delegate and budget calculation.
  8. Hold a final plan discussion with the NDIA delegate before approval.
  9. Ensure that all new framework plans are primarily flexible, rather than primarily stated or restricted supports.
  10. Provide immediate implementation support and a clear early-review pathway.

Meaningful co-design and consultation

SWAN acknowledges the Department’s recognition that the previous consultation needed more detail, accessible information and meaningful engagement. The current consultation process will only rebuild trust if feedback results in visible, meaningful changes. Changes are only reform if they improve the system and process for users, not just the Government’s fiduciary goals.

The Justice and Equity Centre has explained why an obligation to consider co-design is worth little: ‘a Minister would simply be able to affirm they had regard to the importance of co-design in preparing the instrument, without any requirements to actually undertake a meaningful co-design process’. SWAN has raised the same concerns. When the draft support lists consultation was open for 14 days and extended to 21 only after the community pushed, we said then that 12 weeks with fully accessible information is the minimum for people to respond properly. Every Australian Counts has also asked for full exposure drafts, clear information about assessment and budgeting, strong safeguards for people at risk, and ‘meaningful opportunities for co-design, testing and evaluation’ before any large-scale transition. SWAN supports that position.

All exposure drafts should be released together with explanatory documents, practical examples, impact assessments and accessible versions. Consultation must allow enough time for Disabled People’s Organisations, Disabled Persons and Families Organisations, First Nations disability organisations, regional groups and community organisations to discuss the proposals with their members. This requires a minimum of 12 weeks with full accessible materials.

DHDA should publish a detailed response explaining what it heard, what changed, what did not change and why. It should establish ongoing independent oversight with strong representation from people with disability, families, First Nations people, regional communities and relevant disability cohorts.

Public safeguards and data

The NDIA and Department must publish regular, disaggregated data throughout testing and rollout. The data should identify whether the framework is producing (un)intended funding reductions, increased use of restrictions, higher review rates, unequal outcomes or poorer

SWAN Submission on New Framework Planning Rules October 2026 Page | 24

access in regional and remote communities. There should be reports clearly detailing what percentage of participant budgets are flexible, restricted or stated supports.

The Department has not said whether automation will be used. Its own report records that ‘many respondents questioned, however, whether automated decision-making or algorithms will be used in new framework planning’, and that people worried about ‘a reduction or replacement of human decision-making, or decisions being made that are not based on individual circumstances’. People asked the Department to be transparent about ‘automated processes or algorithms used in new framework planning’ and ‘how data is used by the assessment tools’. The Robodebt Royal Commission reported in July 2023 on what happens when an automated process runs without a lawful basis and without disclosure. The rules should make the NDIA disclose every automated part of the process, ban decisions made only by machine, and give people the right to see the information and reasons behind their budget and decisions made in creating their plan.

At a minimum, public reporting should cover:

  • Budget outcomes comparing new framework plans with old framework plans.
  • Funding reductions and increases.
  • Number of and wait times for plan variations, reassessments and replacement support needs assessments, along with data on rejected requests for these processes.
  • Internal review and ART applications, outcomes and timeframes.
  • Percentages and amounts of flexible funding, stated supports, restrictions and funding periods.
  • Unspent funding and documented reasons for under-utilisation.
  • Check-ins performed.
  • Implementation support provided (number of participants supported, type of support etc).
  • Participant satisfaction, safety concerns and unmet need.
  • Outcomes by disability cohort, location, age, gender, cultural background and other relevant intersectional characteristics.
  • Data on debts applied to participants and change of plan management type against participant’s wishes.
  • Preventable hospital admissions of NDIS participants unrelated to medical needs.

Where the data shows unequal or unsafe outcomes, transition must pause for the affected cohort or location until an independent review identifies and implements a remedy.

Automation of planning decisions

The National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Act 2026 received Royal Assent on $20^{\text{th}}$ August 2026 as Act No. 66 of 2026. Schedule 3, Part 2 inserts sections 59B to 59E into the NDIS Act. Section 59B(1) allows the CEO to arrange in writing for computer programs to take administrative action under designated provisions. Section 59B(4) provides that this includes administrative action involving ‘a discretion being exercised’, ‘an evaluative judgement being made’ and ‘a state of mind being formed’. Section 59B(5) treats action taken by the program as action taken by the CEO.

Four provisions are on the list now: Section 59C(1) specifies sections 33, 45, 45A and 45C. The Revised Explanatory Memorandum explains that these cover the grouping of supports under section 33(2E) and the claims and payment process, where the Agency ‘currently processes over 660,000 claims per day’.

SWAN Submission on New Framework Planning Rules October 2026 Page | 25

New framework planning was left off deliberately. The Revised Explanatory Memorandum says: ‘A decision was made not to list provisions relevant to new framework planning, such as sections 32K and 32L. It is clear that the operationalisation of new framework planning will involve at least computer assisted decision making to support a human delegate in complex calculations or application of methodologies. However, more work must be done on establishing the planning process and engaging with states and territories before decisions on whether and how administrative action by a computer-based program could be used in the new framework planning process.’ Sections 22 and 23, which deal with age and residence, were also left off because ‘sufficient work has not yet been done to establish how the end to end system would operate’.

The Minister can add them later without a Bill. Section 59C(2) allows the Minister to specify, by legislative instrument, a provision of the Act, a provision of a legislative instrument made under the Act, or a provision of a legislative instrument made under the Getting the NDIS Back on Track No. 1 Act 2024. Where the action involves evaluative determinations, section 59C(3) requires only that the Minister be satisfied it is appropriate to specify the provision. There is no requirement to consult participants, to publish an impact assessment, or to wait until the assessment tool has been independently validated. The Rules now under consultation are legislative instruments of exactly the kind section 59C(2) reaches.

The Human Technology Institute at the University of Technology Sydney told the Senate Community Affairs Legislation Committee on $29^{\text{th}}$ May 2026 that explicit legislative authorisation for automated systems to apply to discretionary decisions is, to our knowledge, unprecedented in Australian law. It recommended removing section 59C(2), and said that where administrative action is to be automated, ‘this should be provided in primary legislation, not through delegated legislation’. Section 59C(2) was enacted as drafted.

There is also a gap that sits directly under the budget method. The note to section 59B(1) states that statutory authorisation ‘would not generally be needed if a computer program is only being used to assist a decision-maker to make a decision’. The Explanatory Memorandum expects new framework planning to involve ‘at least computer assisted decision making’. A program that calculates a person’s budget and presents the result to a delegate, which they cannot change, therefore falls outside sections 59B to 59E altogether. No standard operating procedure instrument is required under section 59D. No notice to the participant that a computer program was involved is required under section 59E(4). No statement of the arrangement is published, and none of it appears in the Agency’s annual report.

The aged care experience demonstrates why meaningful human judgement must remain available. Support at Home commenced on $1^{\text{st}}$ November 2025. An ABC Four Corners investigation published on $17^{\text{th}}$ August 2026 reported that clinicians could not change the classification algorithm’s funding outcome, even where they considered it inadequate. The investigation also reported that, of about 900 decisions reviewed during the first five months of the algorithm’s use, 207 reviews were successful. An appeal after an inadequate decision is not a substitute for allowing an appropriately qualified person to identify and correct an error before necessary support is denied.

On $2^{\text{nd}}$ July 2026 the Senate passed a bill to reinstate human override in all aged care assessments, co-sponsored by the Australian Greens, the Coalition and Senator David Pocock. The Guardian Australia reported that ‘the algorithm underpinning the tool has frequently under-assessed people and left them without adequate care, while expert

SWAN Submission on New Framework Planning Rules October 2026 Page | 26

assessors were explicitly prohibited from overriding the tool when it made a wrong assessment*, and that about a fifth of people who contested an automated home support funding decision received more funding after human review. The NDIS is about to introduce a standardised assessment tool and a calculated budget into a much larger scheme, with no access to human-only calculated funding budgets, even on appeal to the Administrative Appeals Tribunal.

A tool does not need statutory authorisation to take over a decision in practice. In its February 2026 report to the NSW Parliament, ‘No capacity to allocate’, the NSW Ombudsman found that 47 of 131 senior Department of Communities and Justice staff, 36%, said a structured decision making tool usually makes the risk of significant harm decision, and 32 of 123, 25%, said the same about whether a child is in need of care and protection. Those tools are described as ‘decision support’.

Australian safeguards compared with European law

Australian privacy law will not fill the gap. From 10th December 2026, Australian Privacy Principles 1.7 to 1.9, inserted by the Privacy and Other Legislation Amendment Act 2024, require an entity that has arranged for a computer program to make, or to do something substantially and directly related to making, a decision that could reasonably be expected to significantly affect a person’s rights or interests to say so in its privacy policy. The disclosure is by category: the kinds of personal information used, the kinds of decisions made solely by a program, and the kinds of decisions for which a program does something substantially and directly related to the decision. It gives no right to an explanation of an individual decision, no right to human review, and no limit on what may be automated.

European law would not permit much of this. Article 22(1) of the General Data Protection Regulation gives a person ‘the right not to be subject to a decision based solely on automated processing, including profiling, which produces legal effects concerning him or her or similarly significantly affects him or her’. Where an exception applies, Article 22(3) requires safeguards including ‘at least the right to obtain human intervention on the part of the controller, to express his or her point of view and to contest the decision’. Under Annex III, point 5(a) of the European Union Artificial Intelligence Act, systems used by or on behalf of public authorities ‘to evaluate the eligibility of natural persons for essential public assistance benefits and services, including healthcare services, as well as to grant, reduce, revoke, or reclaim such benefits and services’ are classified as high risk. That classification brings obligations for risk management, data governance, record keeping, technical documentation, human oversight and registration before deployment.

A support needs assessment tool that sets a person’s disability budget would be considered high risk in Europe, and the person assessed would have an enforceable right to human intervention and to contest the outcome. In Australia the equivalent requirements are a written arrangement made by the CEO, a notifiable standard operating procedure instrument, and a line in an annual report. SWAN is not asking the Department to adopt European law. We are pointing out that Australians with disability are being given weaker protection than a European pensioner or jobseeker already has, in a system that decides how much support they get to live.

Participant data

Commonwealth reporting on automation is thin. The Office of the Australian Information Commissioner reported in January 2026 that 4 of 23 agencies, 17%, disclosed their use of automated decision making, while 64 statutes across 12 portfolios authorise it. The earliest

SWAN Submission on New Framework Planning Rules October 2026 Page | 27

provision it identified was section 6A of the Social Security (Administration) Act 1999, introduced in 2001.

Data matching practice has moved out of Parliamentary sight. Auditor-General Report No. 12 of 2025-26, tabled on 9 December 2025, found Services Australia ‘partly effective in managing the privacy of client information’. Services Australia advised the ANAO that it had produced 32 data matching program protocols between 1991 and 2024, of which 13 were published. The agency stopped operating under the Data-matching Program (Assistance and Tax) Act 1990 after 2016-17 and moved to the voluntary guidelines, which the ANAO found reduces transparency and accountability to Parliament and allows more frequent data exchanges than the nine cycles a year the Act permits.

Assessment data does not stay where it is collected. The Guardian Australia reported on 9th September 2026 that NDIA data is being shared with the Australian Criminal Intelligence Commission, which has used Palantir software as part of the Fraud Fusion Taskforce - a 25-agency body established in 2022 and led by the NDIA and Services Australia. The NDIA said that ‘NDIS participant information is only shared with Fraud Fusion Taskforce partners on a restricted basis where potential criminal conduct is suspected to protect participants from fraud’, and has denied using Palantir software itself. SWAN makes no allegation of unlawful conduct. The relevant point for the forthcoming Rules is that a support needs assessment will collect the most detailed personal information the Agency has ever held about a participant, and there is nothing in the drafted Rules that limits what it can later be used for.

Governments have already accepted a stricter standard for disability data elsewhere. The National Disability Data Asset Charter lists as unacceptable identifying people ‘for any purpose, including compliance purposes, and other enforcement purposes’; making decisions about an individual’s access to Government funding, on the basis that ‘the use of National Disability Data Asset data to adjust individuals’ funding will not be allowed’; and ‘use of data in automated decision-making tools for enforcement action by Government’. Those commitments protect a research asset. Nothing equivalent protects the support needs assessment data that will set a person’s budget.

Activity data from inside a person’s home is already being marketed as a funding input. InteliCare sells sensors that ‘monitor movement, door access, meal preparation, and bathroom visits to build a picture of wellbeing’, and an artificial intelligence engine that ‘understands individual routines, establishing a unique baseline for each person’. Its ‘Smarter Funding’ product ‘helps operators optimise NDIS and Aged Care funding by aligning services to verified resident activity, health trends, and support needs’. Attachment 4 of the consultation paper proposes an ‘informal supports offset: based on frequency and durability of informal supports’. Read together, these describe activity recorded in a person’s home being used as evidence about how much funded support they need.

Consent to that is not real if refusing costs money. Many participants live in supported accommodation where the decision to install monitoring is made by a provider, not by the resident. The Rules should state that declining monitoring technology cannot reduce a budget, alter a risk rating or support an inference that a person needs less support, and that data from monitoring technology cannot be used in a support needs assessment without the participant’s separate, revocable and informed consent.

SWAN Submission on New Framework Planning Rules October 2026 Page | 28

Recommendations

SWAN recommends that the Department of Health, Disability and Ageing, National Disability Insurance Agency and Government:

  1. Delay compulsory transition until the complete framework is demonstrably ready. Publish all eight exposure draft rules together with the assessment tools and manuals, complete budget method, explanatory materials and impact assessments. Allow at least 12 weeks of accessible consultation and publish how community feedback changes the final arrangements. Once feedback on the draft rules has been received, the redrafted rules must then be released for further consultation and feedback before commencement of compulsory transition to new framework planning.
  2. Conduct voluntary, independently evaluated end-to-end pilot testing before compulsory implementation. Test the assessment, report checking, budget calculation, plan approval, implementation and review together, with participant feedback and evaluation at every stage. Use appropriately accredited assessors, publish cohort-specific results and protect existing funding during testing. Participants receiving inadequate pilot plans must be able to retain or return to their previous funding arrangements.
  3. Validate the actual modified assessment tools and their intended funding uses across diverse disability cohorts and communities. Assess reliability, consistency between assessors, accessibility, cultural bias and the capacity to identify daily living, capacity-building, maintenance and preventative needs. Do not assume evidence for an earlier tool or another population validates the modified NDIS process.
  4. Require disability-specific assessor competence and safe, accessible participation. Set binding qualification, training, cultural safety, trauma-responsive practice and conflict-of-interest standards. Guarantee participant choice of necessary accommodations, support people, breaks and multiple sessions, and access to in-person assessment regardless of location. Participants must be able to request a different assessor without disadvantage.
  5. Require consideration of existing individual evidence and correction before budgeting. Fund necessary additional evidence, reuse valid reports and explain any rejection. Provide the draft assessment report to participants before delegate review and budget calculation, with at least 14 days to respond, or four weeks where posted. Provide timely, funded replacement assessments by a different assessor chosen by the participant where the original assessment was materially flawed, and a clear pathway to challenge refusal.
  6. Publish understandable individual budget calculations and prevent formulae from overriding need. Disclose inputs, scores, hours, prices, defaults, caps, offsets and adjustments, with reasons for each funding outcome. Prevent default amounts becoming caps, verify any alternative support before deductions and prohibit double-counting. Provide an effective remedy where the calculation does not meet assessed needs.
  7. Remove the proposed informal-support offset. Treat informal assistance as voluntary, finite and separate from necessary funded support. Do not assume family availability, interpret cultural obligations as unlimited care or create pressure for supporters to leave employment. Consider informal assistance only with the participant’s wishes and the supporter’s agreement, where safe and sustainable.

SWAN Submission on New Framework Planning Rules October 2026 Page | 29

  1. Guarantee genuinely flexible budgets and tightly limit conditions where delegates may restrict the flexible budget. Make flexibility across eligible categories the default, including the ability to supplement inadequate stated supports. Limit stated supports to individually justified exceptions. Restrict flexible funding only at the participant’s request or where significant evidence of misuse remains after substantive implementation support, with individual reasons, the least restrictive approach, time limits and review rights.
  2. Fund regional, remote and isolated participants according to actual access costs. Apply transparent, mandatory loadings for travel, workforce shortages, specialist access and coordination, including underserviced locations outside remote classifications. Investigate service barriers and failure of implementation support before drawing conclusions from under-utilisation.
  3. Protect provider choice and self-directed support. Do not mandate a provider unless requested by the participant, compel budget pooling or require participation in shared support provision. Use direct commissioning only as a last resort, with local people with disability involved in co-design and provider selection, consideration of service quality and complaint history, independent evaluation and practical rights to change providers. Fund reasonable direct-employment costs, employment liabilities and assistance to direct support through peer mentoring, supported decision-making, microboards and circles of support.
  4. Set a 12-month default funding period and preserve longer-plan purchasing power. Shorter periods must not be routinely imposed and should reflect participant preference. Any proposed exception for unreasonable overspending must have clear evidence, prior implementation assistance and review rights. Allow rollover between funding periods and payment of eligible supports delivered in an previous funding period where needed. Adjust longer-plan funding for cost increases and provide supportive check-ins, particularly where substantial under-utilisation may indicate unmet need.
  5. Make urgent plan changes and equipment repairs immediately workable. Provide timely variations, interim funding and human escalation for temporary or ongoing changes, without requiring a full plan reassessment or reopening unrelated supports. Allow available flexible funding to cover essential emergency repairs when stated allocations are exhausted, with realistic maintenance funding and clear reimbursement arrangements.
  6. Preserve effective review rights and continuity of essential support. Provide accessible reasons, evidence and pathways to challenge assessment, support category, budget and replacement-assessment decisions. Require internal reviewers to engage with participants unless they opt out, and address every issue raised. Maintain existing necessary funding where a new plan reduces or interrupts support during review or appeal, and provide access to independent assessment and expert evidence.
  7. Base support eligibility on disability-related purpose, safety and value, not product labels. Reject blanket exclusions that remove effective individual supports, including appropriate electronic noise-cancelling headphones. Clarify the boundaries for access modifications, gate automation, personal-care aids and disability-related transport costs. Publish evidence for exclusions and provide timely, reviewable individual exceptions.

SWAN Submission on New Framework Planning Rules October 2026 Page | 30

  1. Retain replacement supports until an effective alternative is available. Protect existing approvals and pending applications, ordinary household products, emerging technology and newly needed supports in old framework plans. Allow proportionate low-risk purchases and trials without unnecessary reports or whole-plan reassessment. Clarify assistive technology categories, and recognise benefits beyond replacing worker hours.
  2. Preserve practical daily living, respite, exercise and health support choices. Allow respite at home, during the day or overnight, including flexible use of respite funding for additional in-home assistance. Preserve meal preparation and delivery assistance, community-based daily tasks, qualified personal training and exercise physiology. Fund appropriately trained workers and proportionate clinical oversight, and retain necessary communication and behaviour support in health and mental health settings.
  3. Prevent gaps when responsibility moves between categories or systems. Require a funded, accessible alternative and working handover before necessary support transfers elsewhere, with interim assistance and a named escalation contact where responsibility is disputed. Clarify the treatment of plan management, support coordination, psychosocial recovery coaching and peer support, without interrupting effective support relationships.
  4. Provide reliable guidance and substantive implementation assistance. Publish accessible resources, practical examples and on-demand explanations, supported by consistent staff training and an official written-advice service. Protect participants who reasonably follow official advice or have not received adequate implementation support to understand changed obligations. Plans must explain their calculations, restrictions and review pathways, and record goals in the participant’s chosen wording.
  5. Establish an individual readiness and supported transition pathway. Allow time to correct records (minimum of 12 weeks), provide preparation and final planning discussions, and fund independent advocacy and navigation assistance at the level required. Provide a reviewable transition decisions and a clear deferral pathway. Do not transition people during crisis or while errors on participant information files remain unresolved. Explain the intended assessment and planning pathway for children and young people, including interaction with foundational supports.
  6. Prohibit fully automated adverse planning decisions and require meaningful human accountability. Apply transparency, explanation, correction and oversight safeguards to computer-assisted decisions as well as automated action. Record inputs, outputs and program versions and provide access to an authorised person able to explain and correct outcomes. Before any expansion of automation into planning or eligibility, require an exposure draft, at least eight weeks of public consultation, independent validation, an impact assessment and Parliamentary scrutiny.
  7. Protect assessment data, consent and the exercise of rights. Publish binding limits on collection, access, retention, sharing and secondary use of sensitive information. Refusal of in-home monitoring must not affect funding or risk ratings, and monitoring data must not inform assessment without separate, informed and revocable consent. Complaints, reviews, advocacy, information requests, contact with MPs and provider changes must never adversely affect funding, plan management type or increase restrictions.

SWAN Submission on New Framework Planning Rules October 2026 Page | 31

  1. Establish independent oversight, public outcome reporting and enforceable pause mechanisms. Report adequacy, unmet need, service access, budget changes, the proportions of flexible, restricted and stated funding, implementation assistance, check-ins, review outcomes, debts, unwanted management changes and relevant safety outcomes. Disaggregate results across disability cohorts and communities.

Conclusion

SWAN supports reform that makes the NDIS fairer, clearer and more responsive. However, consistency must mean meeting individual disability support needs, not funding people to an average. Transparency must mean being able to understand and challenge a decision before an error causes harm. Choice and control must be reflected in usable funding, practical support options and the ability to direct one’s own life.

The proposed framework will determine whether people have the assistance needed to remain safe, communicate, maintain relationships, work, study and participate in their communities. A budget that appears adequate on paper is not adequate if its calculation overlooks support needs, its restrictions prevent sensible use, or the necessary services are unavailable locally. Nor is a budget fair if it depends on informal supporters providing care they cannot safely or sustainably provide — particularly in the context of plans up to five years in duration.

The Department and NDIA must demonstrate that the complete assessment and planning pathway works before requiring participants to use it. This requires independent validation of the actual tools and their funding uses, a full end-to-end pilot with participant feedback, transparent calculations and meaningful opportunities to correct errors. The promised flexible budget must be delivered in practice, with timely responses to changing needs and continued support during disputes.

Participants must also retain control over their providers, support arrangements and personal information. Technology may assist administration, but it must not displace accountable human judgement or effective review. Exercising rights or declining monitoring must never result in less support or additional restrictions.

SWAN urges the Government not to proceed with compulsory transition until the full framework is publicly available, properly consulted on, independently evaluated and demonstrably safe, accurate and equitable. Where testing identifies inadequate budgets or unequal outcomes, the response must be to correct the system - not require participants and families to carry the consequences.

The NDIS should be secured by improving the quality of decision-making, access, flexibility and accountability. It must not be secured by reducing necessary support, transferring responsibility to unpaid carers or making people navigate more complex processes to obtain the assistance they need.

SWAN Submission on New Framework Planning Rules October 2026 Page | 32

os swan

South West Autism Network

Contact

CEO: Nick Avery
Email: info@swanautism.org.au
Web: www.swanautism.org.au

SWAN Submission on New Framework Planning Rules October 2026 Page | 33