We Read the Senate Committee's Final NDIS Report. Here's What Still Concerns Us.

carli215
6 days ago
18 min read

The Senate Community Affairs Legislation Committee tabled its final report on 14 August 2026 into the National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Bill 2026.

The headline is simple. The Committee recommends that the Bill be passed.

We have now read all 164 pages, including the dissenting reports and the additional comments, and compared them against the submission we made to the inquiry ourselves.

The story is a great deal more complicated than that headline suggests.

The Bill in front of the Senate is also not the Bill that was introduced. It passed the House of Representatives with 30 amendments, 18 of them Government amendments and 12 from the crossbench.

Some of those amendments provide real protection. Some concerns have been partly addressed.

Others have not been addressed at all, and several of the questions we put to the Committee still do not appear to have been answered anywhere on the public record.

The short version

If you read nothing else, read this.

The Committee majority recommends the Bill pass, but there is no consensus. The Australian Greens and Senator David Pocock both say it should not pass in its current form. Coalition Senators say serious concerns remain.
Some supports are now protected from ministerial cuts. Daily living, transport, consumables, assistive technology and home modifications can no longer be reduced by support determination.
Social, Civic and Community Participation is not protected. It can still be cut, and so can Capacity Building Daily Activities. This is our single biggest remaining concern, and Senator Pocock's too.
Automated decision-making powers could commence before the post-Robodebt safeguards exist. The Bill gives software the legal power to form what the legislation itself calls a "state of mind".
Nobody has shown what Australia actually saves. The $37.8 billion figure is a saving to the NDIS budget line, not a whole of government calculation.
Government departments and councils billed roughly $475 million from the NDIS in a single year. That has never featured in the public conversation about scheme sustainability.
The Department admitted it did no specific workforce modelling before a Bill that independent research estimates could cost more than 51,000 full time equivalent jobs.
Access restrictions are expected from 2028, leaving roughly 18 months to design, fund, staff and open the foundational services people are expected to move to.

Read our full submission to the inquiry here:

The recommendation to pass is not unanimous

The Committee majority, chaired by Senator Ellie Whiteaker, made a single recommendation: that the Bill be passed. The final report also contains additional comments from Coalition Senators, a dissenting report from the Australian Greens, and a separate dissenting report from Senator David Pocock.

The Greens, led in this inquiry by Senator Jordon Steele-John, say the Bill goes too far, too fast, and should not pass the Parliament.

They say they stand alongside a chorus of disabled people, disability representative organisations, human rights commissioners, allied health professionals, lawyers, advocates, disability service providers, academics and researchers who, in their words, collectively hold a firm view that the Bill should not pass.

Senator David Pocock acknowledges that the extended inquiry and the House amendments improved the Bill and says he supports the general architecture of the reform.

He also says plainly that he cannot support the Bill in its current form. In his words, two matters remain particularly serious:

"the power to impose blunt, indiscriminate reductions to Social, Civic and Community Participation (SCCP) supports, and the breadth of automated decision-making authorised by this Bill ahead of safeguards recommended by the Robodebt Royal Commission."

He makes seven recommendations.

Coalition Senators, whose additional comments are signed by Senator the Hon Anne Ruston, did not issue a dissenting report.

They raise significant concerns about fraud and integrity, participant safety, and what happens to people deemed ineligible for the NDIS or whose support funding is reduced.

They state that ensuring people remain safe from abuse, neglect, violence and exploitation is "non-negotiable".

So yes, the majority recommends passage. That is not the same thing as agreement that this Bill is ready.

Where the four Senate positions landed

Position	Recommendation	Core concern
Committee majority	Pass the Bill	Considers the House amendments and departmental assurances sufficient
Coalition senators	No dissent, additional comments	Fraud and misuse of taxpayer funds inadequately addressed; participant safety; what happens to people found ineligible
Australian Greens	Do not pass	Goes "too far, too fast"; workforce losses; care shifted onto unpaid family and friends
Senator David Pocock	Cannot support in current form	SCCP reductions; automated decision-making commencing ahead of post-Robodebt safeguards

More than 4,500 submissions and pieces of correspondence

The level of engagement with this inquiry was extraordinary. The Committee received over 4,500 submissions and pieces of correspondence.

It held six public hearings, including in Canberra on 30 and 31 July 2026 and in Perth on 6 August 2026, heard from over 80 organisations, and heard the personal stories of over 25 people with lived experience of disability and the NDIS.

Some submissions had not been published by the time the final report was completed. Submissions received by 10 July 2026 will continue to be published after the reporting date and tabled later in the Senate.

Around 250 additional short statements were received.

Those voices matter, and they should matter when Senators decide what happens next.

Some important protections were added

It is only fair to acknowledge where the Bill has genuinely improved, and worth noting that several of these protections came through crossbench amendments rather than the Government's own.

Support determinations can no longer apply reductions to daily living, transport, consumables, assistive technology or home modifications.

The Government has also indicated that employment supports, disability related health supports and necessary continuous 24/7 supports will be protected.

The provisions around "appropriate treatment" were strengthened. Restrictive practices such as forced medication are excluded, and relevant treatment must be available through Medicare, the Pharmaceutical Benefits Scheme or the public health system.

Safeguards were added around attempts to contact participants before a plan is suspended or revoked, and additional transparency requirements were attached to automated decision-making, including publication of a standard operating procedure instrument.

Those changes matter. They do not resolve every problem.

Social, Civic and Community Participation remains our biggest concern

There is one thing we would change about our own submission. We wish we had devoted far more of it to Social, Civic and Community Participation.

Not because we underestimated it. We already knew how essential these supports are.

But the Bill raised so many structural, legal and practical problems at once that our submission concentrated on ministerial powers, automation, access, review rights, fraud enforcement, planning, reassessment and whole of government impacts.

Looking back, SCCP deserved more of that space.

Despite the amendments, support determinations can still apply to Social, Civic and Community Participation and to Capacity Building Daily Activities.

They are what allows a person to:

maintain relationships
leave the home
take part in community life
access education and employment
build independence
reduce isolation
have meaningful relationships with people outside their own household or their disability provider

Senator Pocock's dissent makes the argument we think deserves more attention than anything else in this report. Community participation is itself a safeguard.

In his words, it means

"more people in a participant's life who know them, notice changes in their circumstances and may recognise when something is wrong. Put simply, the more people who are connected to a person and their community, the safer they are. This is true for all of us."

He notes that social isolation and exclusion are known drivers of violence, abuse, neglect and exploitation of people with disability.

The Disability Royal Commission identified the dangers associated with segregation and isolation. The Disability Discrimination Commissioner told the Committee that reducing these supports would leave people in unsafe situations.

Women with Disabilities Australia and other witnesses warned that isolation can allow violence and abuse to remain hidden.

Pocock's first recommendation is that the provisions enabling SCCP reductions through a support determination be removed from the Bill.

His second is that, if that does not happen, the Bill be amended so any reduction to SCCP supports must take account of the participant's individual circumstances, including their safety and their risk of isolation.

That second recommendation is not a radical ask. It is the minimum.

A funding cut does not make the support need disappear

This was central to our own submission. The Government projects $37.8 billion in savings over four years.

A reduction in the NDIS budget is not automatically a saving to Australia.

Our submission argued that the cost simply moves into:

Centrelink
Medicare
public mental health services
hospitals
housing and homelessness services
state services
the justice system
unpaid family care

We asked the Committee to require whole of government cost accounting rather than assessing savings against a single budget line.

The final report contains evidence that supports exactly that concern.

First Peoples Disability Network warned that when appropriate alternatives do not exist, people are not necessarily redirected into another functioning support system.

They can instead be redirected into gaps, including emergency departments, child protection and police, while families absorb what the formal system fails to provide.

The Greens' dissent puts it in the starkest terms, quoting research that the reduction in paid support hours represents

"care that would need to be undertaken by someone else, most likely for no pay and at considerable opportunity cost to the economy"

The underlying need does not disappear because a budget line gets smaller. It only changes who carries it.

Foundational supports need to exist before people are expected to rely on them

We support strong mainstream and foundational disability supports outside the NDIS. The NDIS was never meant to be the only system supporting Australians with disability.

Senator Pocock says the same thing, and then makes the obvious point that follows from it:

"people cannot be prevented from accessing the Scheme on the assumption that another system will support them if that system has not yet been established."

There is an enormous difference between a system that will exist and a system that is funded, staffed, operating and accessible.

The proposed access changes are expected to begin in 2028. That leaves approximately 18 months to identify who will need support outside the NDIS, design and fund those services, and build the workforce to deliver them.

Pocock calls this "a very short runway" for reforms of this scale.

His recommendations here are practical. Changes restricting access should not commence for an affected group until appropriate foundational or mainstream supports for that group are funded, operating and accessible.

And the Federal Government should report quarterly to Parliament and the public on the development of foundational supports, including what will be available, who can access it, and when it starts.

We agree with both. People cannot be moved out of one system on the assumption that another system will catch them while that other system is still being built.

The inquiry heard serious concerns about permanence and treatment

The amendments around "appropriate treatment" are an improvement, but advocates raised real questions about how the provisions will operate in practice.

Mental Illness Fellowship Australia highlighted the difficulty of assessing people with episodic or fluctuating psychosocial disability.

Someone may present as relatively functional during an assessment while being significantly impacted at another point in time.

Deaf Australia raised concerns that Deaf people could feel pressured towards medical interventions such as cochlear implants before Auslan supports are approved.

The Department and the NDIA have said invasive interventions will not be required and that participant choice will remain part of treatment decision-making.

That reassurance matters. But when eligibility for essential disability supports is at stake, participants deserve protections that are clear, enforceable and difficult to reinterpret later.

There is a further point from our own submission that has not gone away. The Bill states that financial circumstances and geographical location do not excuse a participant from the requirement to try appropriate treatment.

In much of regional and rural Australia those services do not exist, and where they do the waitlists run to six or twelve months.

A requirement that ignores financial and geographic reality does not create access. It creates a lottery.

Automated decision-making remains a serious concern

We asked:

What data will be entered into automated systems?
Who enters it?
What training will they receive?
How will an algorithm reach conclusions on evaluative matters?
What happens when the data is wrong?
What happens when someone's circumstances do not fit the algorithm?
How does someone challenge a decision made through a process they cannot see or understand?
What independent oversight exists?

It is worth being precise about what this Bill does. It gives computer programs the legal power to make decisions, exercise discretionary powers, and form what the legislation itself calls a "state of mind" on behalf of the CEO.

The standard operating procedure instruments that govern how the algorithm makes those evaluative decisions are notifiable but not disallowable.

Parliament cannot strike them down.

The inquiry heard very similar concerns from others.

The Human Technology Institute warned that automated systems require effective legal safeguards, accountability and oversight.

The Health Services Union said algorithms should not become the sole determinant of whether someone receives, loses or has support reduced.

MND Australia argued that technology should support clinical judgement rather than replace it. National Legal Aid raised concerns about automated decisions involving discretion and evaluative judgement, including whether adequate review rights would exist.

Senator Pocock identifies a further problem, and states it plainly:

"These powers would commence before the Government has implemented the government-wide safeguards it accepted following the Royal Commission into the Robodebt Scheme, including legislation governing automated decision-making and an independent body to monitor and audit its use."

He accepts that automation has a role in the NDIS, particularly for routine administrative tasks. But he recommends that decisions requiring evaluative judgement, or the formation of a particular state of mind, not commence until that legal framework and independent oversight body are in place.

He also recommends that any person affected by an automated decision have access to internal merits review by a human decision-maker with authority to reconsider it, noting that

"a participant should not have fewer opportunities to challenge a decision merely because it was made by an automated system."

After Robodebt, none of that should be controversial.

A lot of the detail still comes later

The Senate Standing Committee for the Scrutiny of Bills identified significant matters left to delegated legislation, including:

functional capacity
what constitutes appropriate treatment
certain automated administrative actions involving discretion or evaluative judgement
pricing
monitoring and investigation conditions
the percentage reductions applied to particular support groups

Parliament is being asked to approve a framework while some of the most consequential operational detail is settled afterwards, largely by instrument.

Our submission asked for the new planning framework rules to be published in draft for public consultation before the Bill passes.

We still think that is the bare minimum before 760,000 people have the way they access support fundamentally reshaped.

Fraud is real. More powers are not the same thing as better enforcement

We should be clear about our own position here, because it would be easy to misread us. We support strong enforcement.

Providers who exploit participants or steal public money should be investigated and prosecuted, and we are happy to be held to that standard ourselves.

Several of the integrity measures in this Bill, including record retention requirements, claim time limits and conflict of interest rules for plan managers, are overdue and warranted.

Our concern was never whether regulators should have powers. It was whether they use the ones they already have.

The NDIS Quality and Safeguards Commission already had the power to investigate registered and unregistered providers under its own Compliance and Enforcement Policy.

On the figures in our submission, only 0.22 per cent of more than 7,000 fraud reports led to prosecutions. This Bill gives more powers to the same agency.

There is also an obvious problem with treating registration as the answer to fraud.

Registered providers have defrauded the NDIS at significant scale.

Registration did not prevent it. Enforcement would have.

Registration creates visibility, minimum standards and regulatory obligations, and those things are worth having, but it is not a fraud prevention shield.

Our submission also warned that expanding mandatory registration carries a consequence the Government has not addressed.

Smaller providers, many in regional and rural areas where larger organisations do not operate, will exit rather than absorb the compliance cost.

The result is reduced participant choice and reduced service availability in exactly the places that already have the least of both.

The final report does contain evidence supporting stronger registration. The Australian Federal Police said registration could assist prevention, investigations, prosecutions and the removal of bad actors.

But when asked directly whether the AFP expected to see changes emerging from these reforms in relation to organised criminal activity, Acting Assistant Commissioner Donna Parsons of the AFP's Crime Command told the Committee on 6 August 2026:

"No, not at this stage. The environment obviously changes depending on how organised crime networks or others are adjusting within that space, but we don't see the legislation necessarily having an impact on that."

Coalition Senators concluded that fraud and misuse of taxpayer funds remain inadequately addressed.

The question we asked still stands. Before creating new powers, why did the existing ones not produce the outcomes Australians were promised?

The questions from our submission that still appear unanswered

We went back through the final report and compared it against what we submitted. Some of our concerns were reflected in amendments.

Some were raised by other organisations. Several remain, and these are the ones we would most like to see answered before a vote.

Is the evidence the NDIA already demands actually being read?

In February 2025 the NDIA CEO publicly acknowledged that NDIA staff often do not have time to read the medical reports submitted by participants.

Participants are required to fund and provide clinical evidence at every turn.

Before building increasingly complex evidence hierarchies and assessment methodologies, there is a more basic question.

Is the evidence already being provided being read, understood and applied consistently?

What do incorrect decisions actually cost?

Our submission cited the NDIA's own Independent Expert Review Program Evaluation Report, which put the average cost of a 2.5 day hearing at the Administrative Review Tribunal at $29,899 in hearing days alone, before pre-hearing legal costs.

It also cited tribunal reporting indicating that 98 per cent of cases never reach a hearing.

They settle or are overturned before that point.

If decisions are correct, why are they not being defended? And if they are settling because they were wrong, why is the answer legislation that makes them harder to challenge?

Fix the decision-making. Do not simply make it harder to contest.

Who helps participants navigate the system now?

The Local Area Coordinator service is funded at $2.76 billion. Its stated purpose is substantial, ongoing, individualised engagement with every participant.

In practice, across both branches of our organisation, the vast majority of participants only hear from their LAC when their plan end date is approaching.

Support Coordination currently fills much of that gap, and this Bill proposes to cut Support Coordination spending by 30 per cent and move to a commissioned panel model, while plans auto-renew with no human review.

Someone still has to help people navigate a complicated system. Removing one function does not make that work disappear.

What does "arising directly from an impairment" actually mean?

The Bill changes the wording from supports "arising from" an impairment to supports "arising directly from" an impairment.

No guidance has been provided on what "directly" means in practice.

Disability rarely operates in a straight line. An intellectual disability creates communication difficulties.

Communication difficulties create social isolation. Isolation creates anxiety and depression.

Anxiety and depression create sleep disorders. Sleep disorders worsen the underlying disability.

Every link in that chain currently arises from the impairment and is fundable.

For autistic people, anxiety often arises from the interaction between autism and an environment not designed for them.

Under a strict reading of "directly", are those supports still fundable?

Where does "directly" begin and end, and how will an automated planning system be programmed to interpret it?

How can a person be assessed as if they are alone and funded as if they are supported?

This is the contradiction we think deserves the most scrutiny and has received the least.

Part 1 excludes informal supports and family from the functional capacity assessment. Part 6 requires the CEO to consider what family, informal supports and community can provide before approving funded supports.

The assessment says the family does not exist. The funding decision says the family can do it.

The participant is assessed as if alone and funded as if supported.

Both provisions are in the same Bill. The Government has not explained how they coexist.

Where is the whole of government accounting?

This remains the simplest and biggest unanswered question. What does Australia actually save? Not the NDIS. Australia.

Our submission also identified something rarely mentioned in the sustainability debate.

NDIA FOI Document FOI-25/26-0656, covering the top 1,000 providers by direct payments from 1 July 2024 to 1 July 2025, shows government departments, health services, commissions and councils billing approximately $475 million from the NDIS in a single year.

The department that oversees the scheme billed close to $191 million of that.

Those payments may reflect legitimate services. But a scheme being called unaffordable, in a Bill that gives the Minister power to cut what individual participants can access, contains no equivalent mechanism to scrutinise what government itself draws from the same scheme.

If reduced NDIS expenditure increases:

Carer Payment
Disability Support Pension reliance
hospital presentations
mental health expenditure
homelessness expenditure
justice expenditure
child protection involvement
state disability spending
unpaid caring
workforce exits

then all of that belongs on the other side of the ledger.

A government should not call something a saving without showing Australians where the cost goes next.

What happens to the disability workforce

The Greens' dissent cites research by The Australia Institute, in its report The NDIS cuts: A poor reflection of our national character, estimating that the cuts could result in the loss of 51,641 full time equivalent jobs, representing 93,986,644 hours of work.

Those figures are independent modelling, not a Government forecast. But at a hearing on 30 July 2026, the Department of Health, Disability and Ageing acknowledged that it had not undertaken specific modelling of the Bill's impact on the disability support workforce.

It nevertheless accepted that reductions in support budgets could suppress demand, negatively affect providers and disrupt the workforce.

That is significant. The Greens' second recommendation is that the Government undertake comprehensive modelling of the impacts on workforce participation and develop mitigation strategies.

Before reforms of this scale proceed, Parliament should understand the consequences for the workforce expected to deliver disability support.

This is not only about support worker jobs. The inquiry also heard from participants who rely on their supports in order to remain employed themselves.

Again, the other side of the ledger matters.

Senator Pocock adds a further dimension, recommending that the Government undertake and publish a gender impact assessment of the Bill and its implementation, including impacts on women with disability and on people who provide unpaid care.

The Committee has heard. Now the Senate decides

The final report acknowledges the extraordinary contribution made by people with disability, families, carers, advocates, providers, researchers and organisations throughout this inquiry.

The Committee majority recommends the Bill pass. Senator Jordon Steele-John and the Greens say it should not.

Senator David Pocock says he cannot support it in its current form. Coalition Senators say serious concerns remain.

Organisations that raised significant concerns about different parts of these reforms include:

Over 4,500 submissions and pieces of correspondence were received. Those concerns should not become a footnote now that the report has been tabled.

Common questions

Has the NDIS Amendment (Securing the NDIS for Future Generations) Bill 2026 passed?

Not yet. On 14 August 2026 the Senate Community Affairs Legislation Committee recommended that the Bill be passed, but a committee recommendation is not a vote.

The Bill has already passed the House of Representatives with 30 amendments. It still has to pass the Senate, and the Greens and Senator David Pocock have both said it should not pass in its current form.

Will my NDIS plan change straight away?

No. Nothing in this Bill changes an existing plan the day it passes.

The access changes are expected to begin in 2028, and much of the operational detail, including the percentage reductions applied to particular support groups, will be set later through delegated legislation rather than in the Bill itself.

Which supports can still be cut under this Bill?

Support determinations can still apply reductions to Social, Civic and Community Participation and to Capacity Building Daily Activities.

Following amendments, they can no longer apply to daily living, transport, consumables, assistive technology or home modifications.

The Government has also indicated that employment supports, disability related health supports and necessary continuous 24/7 supports will be protected.

What is a support determination?

It is an instrument that allows a set reduction to be applied to a category of funded supports across participants, rather than a decision made about one person's individual circumstances.

That is why Senator Pocock describes the SCCP provisions as permitting "blunt, indiscriminate reductions", and why his fallback recommendation is that any reduction must at least take account of the individual participant's safety and risk of isolation.

Can a computer make decisions about my NDIS plan?

The Bill permits automated administrative action, including action involving discretion or evaluative judgement and the formation of what the legislation calls a "state of mind" on behalf of the CEO.

Senator Pocock recommends that these powers not commence until the government-wide safeguards accepted after the Robodebt Royal Commission are in place, and that anyone affected by an automated decision have access to internal merits review by a human decision-maker.

How do I contact my Senators about the NDIS Bill?

Use the Parliament of Australia list of Senators to find the twelve Senators for your state, or the two for your territory.

They represent you whether or not you voted for them. Put the Bill's name in your subject line so your email is recorded against the right issue.

Make sure Senators hear from you

If you are an NDIS participant, family member, carer, worker, provider or advocate, now is the time to speak.

Tell Senators what these supports mean in real life. Tell them what Social, Civic and Community Participation allows you or someone you love to do.

Tell them whether the foundational services expected to replace NDIS support actually exist where you live. Tell them what happens to your family if funded support becomes unpaid care.

Tell them what happens to your work, study, independence, safety and community connection. And tell them what must change before they vote.

How to do it:

Open the Parliament of Australia list of Senators.
Find the twelve Senators for your state, or the two for your territory. They represent you, and you do not need to have voted for them.
Write a short email. Three or four paragraphs is plenty. One specific real example of what a support does in your life will do more than any amount of policy argument.
Put the Bill's name in the subject line so it is recorded against the right issue: NDIS Amendment (Securing the NDIS for Future Generations) Bill 2026.

Over 4,500 voices contributed to this inquiry. Do not let them become a statistic buried inside a 164-page report.

If you need to talk to someone

None of this is easy reading, and for many people these changes land on top of a great deal of existing stress.

If you are struggling, there is support available and it is free.

Lifeline: 13 11 14. Crisis support and suicide prevention, 24 hours a day, seven days a week. Phone, text and online chat.
Beyond Blue: 1300 22 4636. Support for anxiety, depression and emotional distress.
13YARN: 13 92 76. A 24/7 crisis line answered by Aboriginal and Torres Strait Islander Crisis Supporters.
1800RESPECT: 1800 737 732. 24/7 counselling for sexual assault, domestic and family violence.
Disability Gateway: 1800 643 787. Information and referrals for people with disability, their families and carers.
National Disability Abuse and Neglect Hotline: 1800 880 052. If you or someone you know is being abused, neglected or exploited.

If someone is in immediate danger, call 000.

If what you need is help understanding what these changes could mean for your own plan, your support coordinator, your Local Area Coordinator, or an independent advocate can work through it with you.

The Disability Gateway can connect you with an advocacy service in your area.

This article is based on the final report of the Senate Community Affairs Legislation Committee into the National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Bill 2026, tabled 14 August 2026, together with our own submission to that inquiry.

Quotations are taken from the report and from Committee Hansard as cited within it. Figures attributed to our submission are drawn from the sources cited in that submission, including NDIA FOI Document FOI-25/26-0656, the NDIA Independent Expert Review Program Evaluation Report (October 2023), and the NDIS Quality and Safeguards Commission Compliance and Enforcement Policy.