NDIA Accountability Project

Submission on the NDIS access changes

‹ PrevPage 1 of 7 · Source p. 1Next ›

Department of Health, Disability and Ageing
NDIS Access Branch
By email: NDISAccessConsultation@health.gov.au

28 September 2026

Re: Submission on the NDIS access changes

IN SHORT

We support the direction of the NDIS access changes. Deciding access on what a person can do is fairer than deciding it on a diagnosis, and the Scheme has to be sustainable. Our concern is how the new tests will work for people with psychosocial disability, meaning disability that comes from a mental health condition, who are the people we see every day.

  • The new assessment measures what a person can do with their surroundings and other people’s help taken out (section 9B). It can’t see whether someone can start a task without a prompt, keep going and do it again tomorrow, which for psychosocial disability is usually where the disability is.
  • From 1 January 2028 an impairment is permanent only if the person has had “all appropriate treatment”, including treatment they couldn’t afford or reach (section 25A(2)). Medicare funds ten therapy sessions a year, so people will be refused for treatment they never could get. The Rules can address this (section 25A(4)).
  • New applicants can be refused because another system should support them (section 25B). For adults with psychosocial disability the other systems are thin, and no one should be sent elsewhere without proof there’s a place for them.
  • The paper doesn’t say who will assess, with what tools, or how to challenge a wrong result. Psychosocial assessment needs clinicians trained in mental health, who are already short and are the same clinicians the treatment test relies on.
  • A lighter process for people who are clearly eligible should run on the evidence people already hold, not on a diagnosis list.
  • Everyone already on the NDIS will be reassessed from 2028. The safeguards should be settled now. Start from the file, no cuts until review rights are used up, and a handover for anyone who loses access.

WHO WE ARE

Unbound Minds Psychology is a psychology practice in Sydney. We see children, teenagers and adults, funded through the NDIS, Medicare, workers compensation and privately. Every week we meet people at the edge of the NDIS, after their Medicare sessions have run out or a public mental health team has discharged them. We comment only on psychosocial and neurodevelopmental disability, because that is what we know.

THE CHANGE THE REFORMS MAKE

The reforms do more than remove the lists. The Act now defines functional capacity, and that definition shapes every decision that follows. The paper uses “standardised” to mean two things. One is a standard process. The same evidence, weighed the same way, wherever you live. We support that fully. The other is a standard environment. Section 9B(1)(b) measures capacity with the person’s environmental and personal circumstances excluded as far as possible. A standard process makes decisions consistent. A standard environment can miss what matters for disabilities that show up in context.

Behind that definition sit two more tests from 1 January 2028, one on treatment and one on other systems. Each makes sense on its own. Together they work on the same person, one after the other. An assessment that samples one occasion understates what’s wrong. If the person passes anyway, they’re asked why they haven’t had treatment they couldn’t get. If they pass that too, a new applicant is pointed to a program that may have no place for them. Nobody knows how many people this will exclude.

ON FUNCTIONAL CAPACITY (SECTION 9B)

The new assessment asks one question the same way for everyone. Does this person’s impairment substantially reduce what they can do in communication, social interaction, learning, mobility, self-care and self-management? To make the answer comparable, the Act measures capacity with the person’s circumstances left out (section 9B(1)(b)) and, in general, without help from other people (section 9B(1)(a)). For psychosocial disability, the risk is what a context-free assessment can’t see, because the disability shows up in the circumstances that have been left out. In noise, under social demand, in unstructured time, when nobody prompts you.

A clinic session is the right way to diagnose autism and a poor way to measure function, because one appointment can’t see whether a person can start a task unprompted, keep it up and do it again tomorrow. Section 9B doesn’t require a single appointment. Nothing yet says the design won’t be one.

Section 9B(1)(b) excludes circumstances “as far as possible”, not absolutely, and it doesn’t exclude time, repetition or initiation, which belong to the person, not the setting. Section 9B(3) lets the Rules set the method and list “matters that may, must or must not be taken into account”. The Rules should say capacity is what a person can reliably keep doing over time and across the whole activity, and that variability, prompting and supervision must be taken into account.

Hypothetical example. A 24-year-old woman with autism and severe anxiety sits an assessment in a quiet room with one examiner. She plans a meal, describes a bus route and finishes a problem-solving task. At home she hasn’t cooked without prompting in months, misses appointments unless her mother reminds her that morning, and has left two jobs in their first weeks after sensory overload. Take out her environment and her mother’s prompting, and the assessment records an ability she can’t sustain.

ON PERMANENCE AND TREATMENT (SECTIONS 24(5), 25(1B) AND 25A; QUESTION 6)

The NDIS is for permanent disability. What changes is how permanence is decided. The government’s concern is that people have been entering the Scheme with conditions treatment could have improved. So from 1 January 2028 an impairment isn’t permanent unless the person has had “all appropriate treatment” and it is likely to last for life (sections 24(5) and 25(1B)).

Section 25A(1) defines appropriate treatment as evidence-based treatment that “can reliably be expected to materially improve, reverse, or alleviate the impact of” the impairment. Section 25A(2) then says treatment can be appropriate “regardless of whether the person’s individual circumstances restrict the person from accessing the treatment”, and its note names “financial circumstances and geographical location”. The Explanatory Memorandum says a person’s circumstances “are not relevant” because access to mainstream services “is the responsibility of all mainstream support systems”. This reverses a Federal Court decision, NDIA v Davis [2022] FCA 1002, which had accepted that treatment a person can’t afford isn’t available to them.

We’re not asking the Department to reopen that decision, only to take its own explanation seriously. If getting treatment is the mainstream system’s job, someone should check that it did its job before a person is refused for not having treatment. Otherwise the person is refused twice, once by the system that couldn’t treat them and again by the Scheme because they weren’t treated.

  • What treatment is actually available. On paper, publicly funded treatment for mental health conditions exists. In practice it’s often not available at the dose that works. Better Access funds up to ten individual sessions a calendar year, while NICE recommends at least 16 planned sessions of CBT for psychosis. None of this is the person’s fault, and under section 25A(2) none of it counts.
  • What ongoing treatment means. For many people with permanent psychosocial disability, treatment never ends. Long-acting antipsychotic injections, ongoing therapy and the occasional admission are the plan, not a sign the condition is temporary. Note 2 to sections 24(5) and 25(1B) recognises this, and it needs to work in practice, or “still needs treatment” will be read as “hasn’t finished treatment”.

Hypothetical example. A 38-year-old man has lived with schizoaffective disorder for fifteen years and gets a long-acting antipsychotic injection every few weeks. The treatment he hasn’t had is a full course of psychological therapy for psychosis. Better Access covers ten sessions and he used them by June. Most days he can’t keep a routine, manage money or leave the house alone. Under section 25A(2), the reason he hasn’t had the treatment doesn’t matter.

The Act gives the Rules a way to deal with this. Section 25A(3)(b) says a person is taken to have had all appropriate treatment “in the circumstances determined under subsection (4)”, and section 25A(5) lets the Rules treat different impairments differently. The Explanatory Memorandum says the power exists “to capture circumstances where it may be inappropriate to determine there are treatment options available for some impairments”, with psychosocial disability as its example. The Rules should say a person is taken to have had all appropriate treatment when any of the following applies. If the Rules can’t do this, section 25A(2) should be amended.

  • they sought publicly funded treatment and couldn’t get it within a set time, because of waiting lists, workforce or distance;
  • public funding doesn’t cover the dose that works, and they’ve used what it covers;
  • they’ve had a proper course of the treatment and it didn’t improve what they can do, even if it improved their symptoms;
  • the treatment isn’t safe or suitable for them, in their clinician’s view; or
  • their impairment is a psychosocial impairment of a kind the Rules name, as the Explanatory Memorandum contemplates.

ON OTHER SYSTEMS (SECTIONS 25B AND 34(1)(g))

The NDIS was never meant to be the only system that supports people with disability, but in practice it has often been the only door that opens. From 1 January 2028, section 25B makes the “alternative support requirements” a condition of access for new applicants, section 30(1)(c) lets the Agency remove a participant who joined after that date and stops meeting them, and the Rules can declare a support offered through another system to be an “alternative support” (section 25B(4)). From 1 February 2027, section 34(1)(g) rules out funding, for anyone, a support that “would be more appropriately provided or funded by” another system.

For adults with psychosocial disability, the other side of that boundary is thin. The Commonwealth Psychosocial Support Program is the main program outside the NDIS. The Department’s own 2024 analysis found about 230,000 people with severe mental health conditions going without the psychosocial support they need, and on 1 July 2026 the Minister told the House that those people are “effectively getting no support whatsoever”. The program is built for “short-term help to function day to day”, its services carry waiting lists in our experience, and its current funding ends on 30 June 2027.

NSW funds Community Living Supports, contracted to June 2031, and a person sent to it needs a place in it. Governments have promised $10 billion of Foundational Supports outside the NDIS, but the only part that exists is Thriving Kids, for children aged 8 and under. For adults with psychosocial disability there is nothing designed, funded or dated.

A program with no place for the person can’t be the better system. No support should be declared an alternative under section 25B(4) unless it is funded and running where the person lives, with evidence of available places. No one should be refused or removed for another system unless the decision names the service, shows a place is available to this person, and says who owns the handover. A refusal letter that lists other services has only moved the work to the person least able to do it.

ON HOW ACCESS DECISIONS GET MADE (QUESTIONS 3 AND 4)

The paper asks what a fair process needs and how it can be lighter for people who are clearly eligible. It doesn’t yet say who does the assessment, with what tools, or what happens when someone disagrees with the result. Those questions sank the 2021 Independent Assessments, which moved the evidence from clinicians who knew the person to assessors the Agency chose and paid. That doesn’t by itself make an assessor wrong, but

whoever assesses, the safeguards have to be written down. No access-outcome targets, published methods, evidence from the clinicians who know the person properly considered, assessor conflicts declared, and a practical way to correct an assessment.

  • Standardise the evidence, not the tool. Consistency comes from asking the same questions of the same kinds of evidence and applying the same threshold, not from everyone sitting the same test with a stranger. The framework should say which areas of life are assessed, what evidence counts, over what period, and where the line for “substantially reduced” sits. The person’s own account should come first, with family, carers and clinicians adding to it where self-report understates the problem.
  • Match the assessor to the disability, and count them first. No single profession covers all six areas of life, so psychosocial assessments will fall to psychologists, mental-health-trained occupational therapists, social workers and mental health nurses, all in shortage. The paper doesn’t say how many assessments the process will need, which is every new applicant plus every current participant over three years, or where the assessors will come from. Every assessment draws on the same scarce clinicians the treatment test assumes are available to treat.
  • Make the lighter pathway depend on evidence, not diagnosis. A list of conditions that count as clearly eligible would rebuild the access lists under a new name. Where clinical records already show substantially reduced capacity over time, the Agency should accept them through a standard template from a clinician who knows the person. Those records belong to people who could afford regular care, so for applicants whose records are thin because of cost or access, the cost of gathering evidence shouldn’t fall on the applicant.

ON DIFFERENT GROUPS OF PEOPLE (QUESTION 5)

The paper asks what will make the process work for people with communication, cultural, cognitive, degenerative, fluctuating, episodic or complex needs. Standard tools are built on averages, and these are the people furthest from the average. For the people we see, it means two things.

  • Conditions that come and go. Many mental health conditions run in episodes, and one assessment is one point on that curve. Function should be assessed over a set window, twelve months by default, using the person’s account, the people around them and records of the worst periods, and “substantially reduced” should be met when function over the whole window is substantially reduced, even if there are good weeks.
  • Children, Thriving Kids and the NDIS. From October 2026 a young child with developmental delay or autism might belong in Thriving Kids or in the NDIS, and the paper says the Technical Advisory Group won’t advise on where the Thriving Kids line sits. In our May 2026 submission to the NSW Department of Communities and Justice we said the Thriving Kids threshold should sit alongside a defined NDIS interface. It still isn’t defined. A child should be assessed once, through one pathway, and a child who misses the NDIS line should be enrolled in Thriving Kids by the same decision, where the child meets its criteria and the family agrees.

ON THE GUIDING PRINCIPLES (QUESTIONS 1 AND 2)

The seven draft principles are the right foundation, but as drafted the law can defeat them. Principle 4 promises to recognise “individual circumstances” and Principle 5 a process that works for “fluctuating, episodic or not immediately visible” disability, while section 9B(1)(b) excludes those circumstances as far as possible. None of the seven says who assesses, or on what safeguards. We’d change one thing and remove nothing.

  • Add to Principle 3 the sentence “Functional capacity is assessed as it is reliably sustained over time and across the whole of an activity, not as it is demonstrated on a single occasion.” That gives Principles 4 and 5 a meaning that survives section 9B.

ON TRANSITION AND REASSESSMENT

The new capacity and permanence tests also apply to everyone already on the NDIS, who will be reassessed over three years from 2028 (Schedule 1, items 11 and 94). That is a much larger group than the people who apply each year, and one that already relies on its supports. The paper leaves transition to a later consultation, but this is where the risks above will do most of their work. Three safeguards should be settled before the Rules are made.

  • Start from the file. A participant’s file already holds years of evidence. Reassessment should start there, and only require a new assessment where the file isn’t enough.
  • No cuts before the decision is final. No one should lose funded supports until their review rights on the reassessment decision are used up.
  • Hand over when access is lost. A participant exited under the new tests should leave with a named service, a confirmed referral and someone responsible for the handover, and the Department should publish how many were exited and where they went.

WHAT WE’RE ASKING FOR

  1. Rules under section 9B(3) that measure capacity as what a person can reliably keep doing over time and across the whole activity, with variability, prompting and supervision taken into account, and a standard evidence framework rather than a single tool, tested for each group before 1 January 2028.
  2. An assessor model in the Rules and a guiding principle to match, with assessors qualified in the disability they assess, no access-outcome targets, published methods, conflicts declared, treating clinicians’ evidence weighed, and a right to review the assessment as well as the decision.
  3. The assessment workforce modelled and published before the design is settled.
  4. Rules under section 25A(4) that treat the treatment requirement as met where publicly funded treatment was sought and not available in a set time, isn’t funded at the dose that works, was completed without improving function, or isn’t suitable for the person, and for psychosocial impairments the Rules name. If the Rules can’t do this, amend section 25A(2).
  1. Guidance that gives Note 2 to sections 24(5) and 25(1B) real effect, so that needing ongoing treatment doesn’t stop an impairment being permanent.
  2. No support declared an “alternative support” under section 25B(4) unless it is funded and running where the person lives with places available, and no one refused or removed for another system unless the decision names the service, shows a place is available, and says who owns the handover.
  3. A lighter pathway based on evidence, with existing clinical records accepted through a standard template, the cost of gathering evidence kept off applicants whose records are thin because of cost or access, and information shared with consent between access, the section 32L needs assessment and reassessment.
  4. Children assessed once, through one pathway that sends them to Thriving Kids or the NDIS.
  5. The framework tested on people with psychosocial disability before it starts, access outcomes published by disability type, region and group, and the 2029 statutory review asked to look at misclassification.
  6. Transition safeguards settled now, and consultation on the draft Rules under sections 9B(3), 25A(4) and 25B before they are made.

CLOSING

We’d welcome the chance to contribute to the Technical Advisory Group’s work on psychosocial and neurodevelopmental assessment, and to pilot a standard evidence template with our clinicians. We’re available to meet at any stage. We consent to this submission being published.

Yours sincerely,

Justin Khoury
Chief Executive Officer
Unbound Minds

REFERENCES

Department of Health, Disability and Ageing, Public consultation paper, NDIS Access Changes (August 2026).

NDIS Amendment (Securing the NDIS for Future Generations) Act 2026, Schedule 1, Parts 1, 6, 8 and 9, and Revised Explanatory Memorandum, items 89 and 92; fact sheet on amendments to the Bill (August 2026).

National Disability Insurance Agency v Davis [2022] FCA 1002.

House of Representatives Hansard, 1 July 2026 (Minister for Health, Disability and Ageing).

Department of Health, Disability and Ageing, Analysis of unmet need for psychosocial supports outside of the NDIS (2024); program pages for the Commonwealth Psychosocial Support Program, Foundational Supports and Better Access.

NSW Health, Community Living Supports (2026); NICE guideline CG178, Psychosis and schizophrenia in adults.

Unbound Minds Psychology, submission on the Draft Thriving Kids Program Specifications (NSW Department of Communities and Justice, 20 May 2026).