Kate Chaney (Curtin, Independent) Share this | Hansard source

I rise to speak on the National Disability Insurance Scheme Amendment (Securing the NDIS for Future Generations) Bill 2026. The NDIS is one of Australia's most important social reforms. At its best it changes lives. It enables people with disability to live with dignity and independence and to be genuinely included in their communities. It allows carers to participate in work and public life.

The economic and social cases for a well-functioning NDIS are not in dispute, and I want to say that clearly at the outset because this is a contested bill and I do not want the contention to be mistaken for opposition to the scheme itself. The people who engaged with me most intensely in preparing for this debate—participants, carers, clinicians, advocates—are not opponents of a sustainable NDIS; they are its most passionate defenders. But they're frightened of what this bill might do to the scheme on which they rely.

I support the need for reform. The NDIS has grown faster than projected. Some of that growth reflects genuine unmet need that was always going to emerge as the scheme matured, but some growth reflects poor initial planning decisions, provider overcharging and inadequate fraud prevention. Reform directed at those drivers is necessary and overdue. The question this bill must answer is not whether to reform; it's whether this reform, as designed, will make the scheme more sustainable without removing genuine support from people who need it to live full lives.

I do not believe the bill as currently drafted passes that test. I'll explain why, drawing on the extensive community engagement I've conducted since these changes were announced on 22 April. Following the government's announcement, I heard immediately from Curtin constituents who were worried—parents of children with disability, adults managing complex permanent conditions, carers who've built their lives around a support system that for the first time recognised their family member as a full citizen with the right to genuine support.

I wanted to listen properly, so I convened a community workshop on 4 May attended by 70 people—NDIS participants and their families, occupational therapists, speech pathologists, physiotherapists, plan managers, support coordinators, providers and independent advocates. Many had been navigating this scheme for years. They came prepared with detailed knowledge and specific practical concerns. I followed that with an online survey, which received more than 90 responses, and an invitation for written submissions. Many of those who wrote did so at length, providing carefully argued accounts of how these proposed changes might affect their lives. Across all channels. I've considered almost 200 responses from my community.

What I heard was not a rejection of reform. Almost everyone accepted that the NDIS must be financially sustainable. What they're asking for with urgency and consistency is that reform be carefully co-designed with people who have lived experience and directed at the real drivers of cost growth—fraud, poor initial planning and provider overcharging—rather than at participants themselves.

Three themes were consistent across virtually every piece of feedback I received. The first is that sustainability must be pursued through efficiency, not exclusion. The people I heard from do not accept that a sustainable NDIS requires removing people who genuinely need support. Savings should come from eliminating fraud, reducing bureaucratic overheads, improving the quality of initial planning decisions and addressing provider overcharging.

The second is that reform must not cost shift to other systems that are not prepared to bear additional responsibility. Clinicians with direct experience in the justice, health and education systems stressed that reducing NDIS supports does not eliminate need. It transfers costs to hospitals, schools, mental health services, aged care and the justice system, often at far greater public expense.

The third is that the critical details, including the thresholds that will determine eligibility under the new functional assessment framework, are yet to be set by subordinate legislation that does not yet exist. Community engagement will happen after the legislative architecture is already in place, and that seems to be in the wrong order.

I will go through the key concerns schedule by schedule. First is the functional capacity assessment. Of all the elements of this bill, the new functional capacity assessment framework generated the most concern and the most detailed feedback from my community. I want to convey the depth and sophistication of what I heard, because it goes well beyond 'we don't want change'. My constituents raised serious, well-grounded concerns about whether a standardised assessment tool can adequately capture the full diversity and complexity of disability, including episodic and fluctuating conditions where a snapshot assessment may not reflect genuine support needs; including masking, particularly for autistic individuals who appear more capable in an assessment setting than they are in daily life; including rare and complex conditions where assessors may lack the specialised knowledge to correctly identify need; and including children, where different tools and developmental frameworks are required.

Allied health professionals raised specific concerns about assessor training. Occupational therapists complete four years of professional training. Underqualified assessment risks systematically underidentifying need, with potentially serious consequences for some of the most vulnerable people in our community.

The concern I heard about most consistently, raised with explicit historical justification, was around the new opportunities for automated decision-making proposed in this bill. Alan, a disability advocate, told me he'd already seen in robodebt what happens when inadequate human oversight of algorithmically driven processes is in place, and he's worried we're about to see a repeat in disability. The community has specific, recent, evidence based reasons to be alarmed, and I'll address this directly in my second reading amendment.

Second, I will look at social and community participation funding. The proposed ministerial determination to reduce social, civic and community participation funding generated a strong response from my constituents. This funding is not discretionary spending in any meaningful sense. For many participants, it's what makes the difference between a full life and an isolated one. It funds the sporting activities, outdoor experiences and community engagement that allow people with disability to be members of their community, to build skills and independence and to experience the ordinary dignity of participation in public life. The $200 million Inclusive Communities Fund, proposed as a partial replacement, was widely viewed as a step backwards towards the segregated, group based models that the NDIS was specifically designed to move beyond. A constituent, Bianca, told me she's worried the proposed inclusive communities program will result in corralling disabled people together into groups where they can only mix amongst themselves.

My constituents have also noted with alarm the sequencing. Cuts are scheduled from October 2026, while foundational supports are not yet operational, not yet funded in many jurisdictions and not yet demonstrated to deliver genuine community inclusion. That's not a transition plan. That's a gap.

On fraud and integrity, my community strongly supports action on fraud. There was no ambiguity on this point. The frustration I heard was not with the goal of fraud prevention but with whether the measures are appropriately targeted. Allied health professionals are already regulated through APRA. Additional NDIS registration can cost upwards of $10,000 and may drive small, high-quality independent providers out of the scheme entirely, leaving participants with only large corporate providers. My community supports a risk tiered approach: higher requirements for closed, high-risk settings, and proportionate, lighter requirements for professionals already subject to rigorous professional regulation.

The next concerns were around plan management and pricing. The proposal to commission a panel of plan management providers, replacing the current open market of over 1,400 plan managers, was contested. I received feedback that the panel model risks eliminating choice and control, consolidating the market in favour of large providers and removing the small, specialist plan managers most valued by participants with complex needs.

I've submitted the detailed feedback I received from my community, together with 22 formal recommendations for improving this bill, to the Senate Community Affairs Legislation Committee inquiry, and I urge that committee to consider these asks carefully.

I now move the amendment circulated in my name:

That all words after "notes that" be omitted with a view to substituting the following words:

"(a) the bill enables the use of automated decision-making in determining a wide range of matters affecting NDIS participants;

(b) it is reported that neither human decision-makers nor the Administrative Review Tribunal will be authorised to modify or override the automated decisions made about the support needs of NDIS participants;

(c) when used well, automated decision-making can deliver faster, more consistent and more efficient government services, and that agencies like the NDIA need digital tools to operate at the scale required;

(d) the Government has failed to implement the recommendations of the Royal Commission into the Robodebt Scheme, handed down in July 2023, which called for the introduction of a consistent legal framework and oversight for automated decision-making in government services;

(e) the consequences of poorly implemented automation for high-risk decisions are already being seen in widespread concern around automated aged care assessments and the unlawful cancellation of income supports for Australians under the Targeted Compliance Framework; and

(2) calls on the Government to bring forward a legislative and mandatory framework for automated decision-making in government services, incorporating transparency requirements, decision-level safeguards, human accountability and oversight for high-risk decisions, meaningful review rights, and independent oversight, in order to prevent a repeat of the failures of the Robodebt scheme".

This amendment addresses the specific concerns raised by my community about the risks of automated decision-making proposed in this bill. Schedule 3 gives the NDIA CEO explicit legal authority to automate a wide range of administrative decisions about what goes into a participant's plan, whether a claim gets paid and whether the payment falls within the pricing limit. The bill allows the minister to extend the scope of automation further through legislative instrument. In other words, this bill gives the government the power to automate almost any administrative NDIS decision, including, as the explanatory memorandum tells us, the new framework planning process, which will determine every participant's total support budget.

The bill does include some transparency measures. Standard operating procedure instruments must be published on the federal register, and the CEO must disclose which provisions have been automated and report annually on cases where automated decisions were overridden. These are steps in the right direction, but the safeguards are not adequate to the risk. We've been told that NDIA staff will have no authority to modify a computer generated budget, even where they identify an error, and, if a participant appeals to the Administrative Review Tribunal, the tribunal cannot remake the decision. It can only return the participant to the same automated system that got it wrong in the first place. This is not a hypothetical risk. The Commonwealth Ombudsman found last year that automated processes in the Department of Employment and Workplace Relations illegally cancelled the income support of nearly a thousand Australians. In aged care, automated assessment tools are generating incorrect support outcomes, and the lack of transparency about how those tools operate is making it nearly impossible to identify and fix systemic errors.

This is a government that's expanding automated decision-making across portfolios—the NDIS, aged care, social security, environmental approvals—without having implemented the consistent legislative framework that its own royal commission said was necessary. In July 2023, the Royal Commission into the Robodebt Scheme recommended that the government legislate a consistent legal framework and oversight mechanism for automated decision-making in government services. The government accepted those recommendations. Consultation followed. Nothing has been legislated. Meanwhile, the use of automation expands.

Robodebt was unlawful. It harmed hundreds of thousands of Australians, and it persisted because there was no independent body with the power and resources to identify and stop it. My amendment calls on the government to legislate that framework with transparency requirements, decision-level safeguards, genuine human oversight for high-risk decisions, meaningful review rights and an empowered independent watchdog. Automated decision-making done well can make government faster and more consistent, but it needs guardrails. Right now, the government is expanding automation across the lives of some of Australia's most vulnerable people without them. Australians with disability deserve decisions made with care, not decisions made by algorithms with no-one accountable for the outcome.

In consideration in detail, I'll be supporting a number of amendments from members of the crossbench that relate directly to other concerns raised by my community. The amendment requiring functional capacity assessments to have regard to environmental, social and personal circumstances restores the contextual dimension to the bill that current drafting would remove, directly addressing concerns about masking episodic conditions and the limits of standardised tools. The amendment requiring ministerial reporting on foundational supports readiness before the relevant parts commence addresses my recommendation that cuts not proceed before replacement services are demonstrably operational. The amendments making support determinations reviewable decisions, requiring parliamentary approval for material funding reductions and strengthening safeguards in section 34A, address my community's concerns about accountability and appeal rights. The amendment requiring genuine consultation before significant NDIS rules commence and an independent statutory review after 12 months reflect my recommendations on co-design and governance. Taken together, these amendments represent what responsible crossbench engagement looks like—not blocking a reform this scheme needs but insisting it be done in a way that does not harm the people it is meant to serve.

The people from my community will be watching—the woman in Churchlands in her late 60s who is caring for her non-verbal son alongside a husband with cancer, who told me that, without his funding for community outings, he would simply sit at home becoming agitated and distressed; the mother in Wembley Downs whose son masks exceptionally well, appearing articulate and compliant but dysregulated at home; and the parent of a non-verbal son with complex behaviours. I urge the government to engage seriously with the 22 detailed recommendations put forward in my community submission to the Senate committee. They come directly from the people this legislation will affect and are practical, considered and offered in good faith.

See this speech in context