Zali Steggall (Warringah, Independent) Share this | Hansard source

The NDIS must be made fiscally sustainable, but sustainability cannot come at the cost of Australians with disability being left worse off. The National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Bill 2026 contains reforms that are genuinely necessary—stronger fraud controls, provider accountability and modernised governance—and I support those measures. But this bill also contains reforms that are moving too fast without the supporting infrastructure, independent modelling or accountability frameworks that responsible reform demands.

Warringah has heard from participants, carers, occupational therapists and providers, and their message is consistent: fix the system, don't punish the people it exists to serve—NDIS participants and their families, many living with autism, psychosocial disability, Down syndrome and complex fluctuating conditions. My electorate office has heard from local NDIS providers, including Northside and Bushlink, one of Warringah longest serving community disability organisations; allied health professionals—occupational therapists, speech pathologists and social workers, all of whom work with NDIS participants daily; and parents of children with developmental delay and autism, who are anxious about the Thriving Kids transition. The NDIS is essential to each of their lives. Constituents are reporting change-of-circumstance requests sitting unprocessed for months, and some have reported drastic funding reductions—in one case, from 36 hours of support to just seven, without explanation or clinical justification. Families of young children are deeply anxious about the Thriving Kids transition, uncertain as to whether community services will actually exist in Warringah when their child's NDIS plan ends. Many of these constituents have raised their view that this bill appears to be driven by a cost reduction agenda, not by ensuring people receive reasonable and necessary supports

In 2022-23, NDIS expenditure grew by 20 per cent. The government committed to an eight per cent growth cap by July 2026, but the latest figure shows that it is still exceeding that. There are currently 774,456 NDIS participants. Without reform, that figure is projected to be exceeded, and that is an issue. In 2026-27, the budget projects some $37.8 billion in reduced expenditure growth over four years. The NDIS spending on core social and community participation reached $11.9 billion in the 12 months to 31 March 2026, and the government says average actual spend in that category has risen from around $14,000 five years ago to about $31,000 this year. I agree: these are real sustainability pressures that cannot be ignored.

But the system's existing failures are also real and can't be dismissed. Participants in my electorate face chronic delays, inconsistent decisions, poor plan quality and an overwhelmed appeals system, with a backlog estimated at 35,000 change-of-circumstance cases. Non-clinically-trained planners are making consequential decisions about complex disabilities. Significant public funds are being spent on administration, tribunals and consultants—not supports.

Fraud and provider misconduct are genuine issues, but this claim is often conflated with legitimate participant need to justify eligibility tightening. It's concerning that there are even allegations that there was messaging consultation done by the government that actually said, 'If you amplify the concern around fraud, you will get more social licence for reining it in and for these changes.' In some ways, it's almost like it was a marketing strategy to enable these changes to be pursued. The government should absolutely pursue fraud, but that cannot be the whole story. The bigger issue is that the scheme has been allowed to operate with insufficient oversight. The regulator does not have full visibility over the provider market, creating space for overcharging, poor practice and exploitation to take hold.

But let's be clear. This is not the only area where it happens. Medicare fraud has been estimated at around $3 billion a year, yet we don't see the Prime Minister and members of the government—when they come in here brandishing their Medicare cards—also concerned about fraud in the Medicare system. We certainly don't see them come in here demonising patients and blaming them for that fraud. We don't describe Medicare fraud as a reason to demonise patients, yet the NDIS is too often discussed as though participants themselves are the problem.

A fair reform agenda would strengthen regulation, close loopholes in the provider market and protect the people the scheme exists to serve. The scheme's integrity problems should not be treated as though they sit with participants. People with disability are not the ones designing pricing models, running provider businesses or exploiting gaps in oversight. Reform should be aimed squarely at those who profit from failure.

This bill gets a few things right. Strengthening fraud controls, civil penalties, expanded information-gathering powers and record-keeping requirements—these are necessary and long overdue. Expanding mandatory registration to higher risk providers—personal care, daily living supports, closed settings—is sensible and consistent with the NDIS Review's risk-proportionate model. But, again, there is overreach in the government's response. The ministerial pricing mechanism provides a lever for cost containment that may be necessary, but there is no precaution or arm's length and the government may not always use this responsibly. There are serious concerns.

Firstly, the I-CAN automated assessment system—transparency and accountability are sorely missing. This question was put to the government today in question time, and there is no reasonable or real answer to this problem. From April 2027, participant support plans will be determined through I-CAN—the Instrument for the Classification and Assessment of Support Needs—a semistructured assessment rating support needs across 12 domains, including communication and mobility. Our current reporting indicates that, once I-CAN generates a budget—and this is an automated system—NDIA staff will have no authority to change it. It will not be reviewed or managed by people. This is an automated system. They can request a reassessment using different inputs, but it is still an automated output. This risks allowing a standardised assessment and budget-setting method to carry decisive weight in determining a participant's support with limited transparency about how the generated budget can be reviewed, varied or overridden.

The assessors conducting these conversations are not required to have allied health qualifications. It's just staggering that we're still at the point of saying that. It's highly desirable but not mandatory. It's deeply concerning. As NDIS expert Dr Georgia van Toorn has noted:

If algorithms are going to determine who gets support and who goes without, then the entire apparatus—including the algorithm itself, its modelling, classification rules and training data—must be open to scrutiny.

The bill permits automated administrative decision-making but does not establish the transparency, the override mechanism or the accountability frameworks needed to make this safe. I'm calling on the government to publish the I-CAN algorithm and modelling, require qualified allied health assessors, guarantee that human override of automated plans is possible and establish an independent audit function for algorithmic decisions. We've seen what happens when automated systems go wrong: they have dire consequences. What the government appears to be doing is repeating the same mistakes. It has done so with aged care. It is repeating the mistakes and ignoring the recommendations that came out of the robodebt royal commission.

Secondly, foundational supports must exist before eligibility is tightened. The bill stages these changes unevenly. Access changes based on functional capacity are not due to commence until 1 January 2028, but changes affecting social and community participation are being advanced before the Inclusive Communities Fund and broader foundational supports are fully designed, funded and actually operating in local communities. We can all see the gap. The $10 billion national agreement on foundational supports over five years sounds substantial, but $4 billion is already committed to the Thriving Kids program for children under nine, leaving just $6 billion for adults and the broader population.

The government has allocated only $200 million over three years for the Inclusive Communities Fund. Just think about that—$200 million over three years to replace social and community participation supports for over 368,000 participants currently using those supports. The maths just don't add up. This sequencing is precisely what the 2023 NDIS independent review warned against: reforms being implemented piecemeal without the supporting structures required. As put to me by a Warringah constituent, rushing these reforms risks 'entrenching cost driven decision-making that undermines both the integrity of the scheme and Australia's disability rights obligations.'

Thirdly, the Thriving Kids transition must have a no-service-gap guarantee. Children with mild to moderate autism and developmental delay under the age of nine will be transitioned from the NDIS to the new Thriving Kids program, which is set to begin rolling out from 1 October 2026. But the NDIS eligibility changes don't take effect until January 2028, meaning there's a real risk that there's going to be a gap between when NDIS support winds back and when Thriving Kids services are available in communities in Warringah and across Australia. We've raised this issue with the government but have yet to be absolutely assured that a service gap will not occur—keeping in mind that these are young children, where every month and period of development is crucial to ensure they are reaching their full potential. State based delivery means geographic inconsistency. Services in Warringah may look very different from those in regional New South Wales or other states. Families have already reported receiving exit letters from the NDIA, claiming children have 'achieved their goals', even when therapists and parents know they still need support. I'm calling on the government to provide a clear, legally binding guarantee that no child will lose access to funded supports until Thriving Kids is operational in their community—not just nationally but locally.

Fourthly, there is no accountability framework for systemic failures. The bill does not address the systemic failures that are driving costs and participant distress—inconsistent decision-making, delays, poor plan quality and a high-volume appeals backlog. Without statutory timeframes for decisions, mandatory clinical input in planning and independent oversight of the NDIA, the structural problems remain and reform risks layering cost cuts on top of dysfunction.

At the consideration in detail stage, I'll be moving three amendments to this bill. The first is about broadening the reassessment trigger. Currently, the bill limits when a participant can request an unscheduled plan reassessment, tying it to a narrow set of circumstances defined by the NDIA. The bill narrows when participants can request an unscheduled plan reassessment, and this amendment adds three practical triggers: when supports in a plan are no longer available, when a plan is not meeting a participant's needs in practice or when new clinical evidence has emerged. These things are not static; they fluctuate. Disability is not static; it needs change. Providers exit markets, and circumstances shift between scheduled reviews.

The minister's power to reduce funding across classes of supports is broad. It is too broad as currently drafted. The bill does not require the minister to consider what that reduction will actually mean for individual participants before acting. My second amendment changes that. Before reducing funding across a class of supports, the minister must actively consider whether the determination would undermine supports assessed as reasonable and necessary for individual participants. This introduces a genuine, holistic safeguard into a ministerial power that currently operates with very limited constraints. It anchors cost containment decisions in the scheme's founding principle—that supports exist to enable Australians with disability to live independently and contribute on their own terms.

My third amendment makes support determinations reviewable decisions, and, in the bill as drafted, a ministerial determination reducing support funding is a legislative instrument, meaning it applies broadly and is not, itself, a reviewable decision that an individual participant can challenge through a merits review. This amendment closes that gap. Where a determination reduces a participant's funding, they are deemed to be affected by a reviewable decision and can access existing review pathways. It does not prevent the minister from acting, but it creates accountability. I want this scheme to succeed; it is so vital for so many. I want it to be here for Warringah families in 10 years and for every Australian with disability who needs it, but I will not support a bill that achieves fiscal sustainability by shifting the burden onto people with disability, their families and their carers. The government's urgency appears to be driven by fiscal and political expediency rather than best care for the most vulnerable.

See this speech in context