Risk of regression in disability support due to tightening access (Participant experience)

Submission 2773

Submission to the Senate Community Affairs Legislation Committee

SUBMISSION TO THE

Senate Community Affairs Legislation Committee

Inquiry into National Disability Insurance Scheme Amendment (Securing

the NDIS for Future Generations) Bill 2026

Personal Submission

Parent of an NDIS Participant | Friend of NDIS Participants

Registered NDIS Provider | 16 Years Experience in Allied Health and Disability Services

May 2026

1. Introduction and Submission Context This submission is made in a personal capacity as a parent of an NDIS participant, as a friend of multiple NDIS participants, and as a registered NDIS provider with 16 years of experience in allied health and disability services. It draws on direct lived experience, professional practice, and detailed observation of the NDIS system and its reforms.

I am not approaching these issues from an ideological or theoretical position. I am approaching them as someone who has worked inside disability systems for over a decade and a half, who has witnessed the harms that arise when systems fail, and who understands — with the clarity that only lived experience can provide — what is at stake for disabled Australians, their families, and the workers who support them.

This submission addresses three interconnected concerns:

1. The historical context of disability support in Australia and the risk of regression

2. The government’s own participant exit modelling and its projected human and economic consequences

3. The risks of automated and semi-automated decision-making within the NDIS, including PACE

4. Historical Context: The Case Against Regression

Submission 2773

Submission to the Senate Community Affairs Legislation Committee

Prior to the NDIS, disabled Australians faced institutionalisation, segregation, chronic waitlists, and denial of support unless families had already reached crisis point. Parents became default full-time carers by necessity, while disabled people were routinely denied access to education, employment, relationships and safety. Governments rationed support and called it sustainability.

Australia’s Disability Royal Commission documented the catastrophic consequences of those systems — violence, abuse, neglect and exploitation at scale. Despite this, the Federal Government accepted only 13 of 222 recommendations in full, with the majority deferred or accepted only in principle.

Current reforms risk repeating this history. Rebranding initiatives such as “Thriving Kids” and “foundational supports” obscure a tightening of access to the individualised supports that keep disabled people safe, stable and included. Proposed approaches that emphasise compliance over wellbeing, require children to normalise to access support, or redirect families to deficit-focused parenting programs are inconsistent with contemporary human rights standards and evidence-based practice.

Disabled people and families are not approaching these reforms from an abstract position. They have lived experience of what came before, and they are clear: these changes represent regression, not reform.

2.1 The Failure of Co-Design Claims of co-design ring hollow when Disabled People’s Organisations are marginalised from the process. Meaningful co-design requires genuine participation — not consultation that proceeds regardless of community objection. You cannot claim “nothing about us without us” while freezing out the very people these changes will affect. That is not co-design. That is managed optics.

Australia’s obligations under the United Nations Convention on the Rights of Persons with Disabilities (UNCRPD) require genuine participation, not the appearance of it.

3. Participant Exit Modelling and the Gendered Cost of Support

Withdrawal

The government’s own Senate modelling projects a reduction in NDIS participants from 944,000 to 598,000 by 2030–31 — representing an assumed exit of 241,000 people from the scheme. This cannot be attributed to improved outcomes. Disability does not resolve itself. The need remains; it is the support that is being removed.

The mechanisms driving these projections are explicit within the modelling:

• Tighter eligibility criteria
• Stricter reassessments
• Ending plan rollovers
• Restrictions on supports
• Changes to children’s access
• A new “substantially reduced functional capacity” threshold

Submission 2773

Submission to the Senate Community Affairs Legislation Committee

Taken together, these represent systemic exclusion by design, not a reform agenda grounded in participant wellbeing.

3.1 The Gendered Impact of Support Withdrawal

When formal support is withdrawn, care does not disappear — it is transferred. The evidence is consistent: that transfer falls disproportionately on women. The downstream consequences include workforce exit, loss of superannuation, financial dependence, and carer burnout.

The broader system costs that follow are predictable and well-documented:

• Increased mental health presentations and hospital demand
• Rising child protection notifications
• Increased family violence
• Children disengaging from school
• Increased homelessness and social isolation among disabled people The framing of disabled people as a budget problem also obscures a significant economic reality: the unpaid labour already being performed by family carers — overwhelmingly women — represents a substantial and unmeasured contribution that existing cost analyses fail to account for.

The NDIS was created precisely because mainstream systems had already failed disabled Australians. Policies that progressively exclude people from the scheme do not resolve that failure — they replicate it.

4. Automated Decision-Making, PACE, and the Limits of “Human in the Loop”

The disability community is experiencing serious and observable harms arising from automated and automation-supported decision-making within the NDIS. These concerns extend beyond fully automated systems to include any environment where human discretion is materially constrained by workflow design, KPI pressures, template-based pathways, and system-generated recommendations.

Review of participant decision letters reveals consistent and troubling patterns:

• Formulaic reasoning and standardised decision language
• Apparent failure to engage with individual evidence
• Outcomes difficult to reconcile with the material before the Agency
• Inconsistent outcomes between comparable participants The high rate of NDIA decisions subsequently varied, remitted, or set aside by the Administrative Review Tribunal suggests these are not isolated errors but systemic failures.

4.1 Why “Human in the Loop” Is Insufficient The government’s reliance on human oversight as a sufficient safeguard misunderstands the problem. International research is unambiguous: humans

Submission 2773

Submission to the Senate Community Affairs Legislation Committee

operating within constrained administrative environments routinely defer to system generated outputs. Nominal review does not restore genuine discretion (Alon-Barkat & Busuioc, 2023; Laux & Ruschemeier, 2025).

Automated systems also carry particular risks in disability contexts, where presentations are complex, fluctuating, and highly individual. Systems optimised for throughput and fiscal targets are structurally ill-equipped to account for nuance, cumulative disadvantage, or atypical need.

When flawed assumptions are embedded in system architecture, they are replicated across entire populations rapidly and consistently. The NSW Child Protection Structured Decision Making model provides a recent domestic example: it was decommissioned after producing racially discriminatory outcomes at scale, increasing rather than decreasing removal rates from Aboriginal families.

The Robodebt Royal Commission established the human and institutional cost of allowing automated systems to outpace lawful governance. The same structural risks are present within the NDIS today.

5. Alternative Funding Considerations Of course, the NDIS must be sustainable. Before accepting that cuts to disabled Australians are fiscally necessary, the government must be asked to account for what it continues to spend subsidising some of Australia’s most profitable industries.

5.1 Fossil Fuel Subsidies

The Fuel Tax Credits Scheme (FTCS) is the single largest fossil fuel subsidy in Australia. In 2024–25 it is projected to cost the federal government $10.2 billion. In total, Australian state and federal governments provided $14.9 billion in fossil fuel subsidies in 2024–25 — a figure that has grown year on year.

Of the $10.2 billion in fuel tax credits projected for 2024–25:

• $4.8 billion flows to the mining industry

• $1.4 billion goes to the coal sector alone

• BHP alone received $622 million in fuel tax credits in 2024 while posting approximately $10 billion in profit and generating more than $78 billion in revenue

• The top 18 mining beneficiaries collectively received $3.36 billion in credits — with the average coal mining company claiming $26 million per year compared to $13,000 for the average farmer

According to The Australia Institute, removing the fuel tax credit exemption for the mining industry alone — while fully protecting farmers — would return $4.8 billion to the budget annually.

5.2 The Fiscal Comparison

To contextualise the government’s stated fiscal position:

• Projected NDIS “savings” from reforms: $19.3 billion over four years

Submission 2773

Submission to the Senate Community Affairs Legislation Committee

• Mining industry fuel tax credits alone: $4.8 billion per year ($19.2 billion over four years)

• Total annual fossil fuel subsidies: $14.9 billion per year ($59.6 billion over four years)

The government could recover equivalent savings through a single industry subsidy reform — without removing one disabled Australian from support. This is a political choice, not a fiscal necessity.

5.3 The Economic Contribution of the NDIS The framing of the NDIS as purely a cost misrepresents what the scheme actually generates economically.

• NDIS spending reached $48.5 billion in 2024–25, representing 1.7% of

Australia’s GDP

• Per Capita calculates the scheme’s total economic benefit at upward of $50 billion per year, including $20+ billion in indirect economic activity

• The Australian Federation of Disability Organisations estimates a conservative economic multiplier of 2.25–3x per dollar invested

• The scheme directly employs over 270,000 people across more than 20 occupations

• Every $1 billion in underfunding produces significant negative economic consequences including direct job losses and reduced community economic activity

The NDIS is not simply an expenditure. It is economic infrastructure — generating employment, stimulating local economies, reducing reliance on crisis services, and enabling workforce participation by both participants and their families.

6. Recommendations

7. Halt projected participant exit targets pending an independent human rights impact assessment, with full publication of the modelling assumptions underpinning current projections.

8. Commission an independent audit of PACE decision quality, examining patterns of formulaic reasoning, failure to engage with individual evidence, and rates of Administrative Review Tribunal variation.

9. Establish an Independent Taskforce on Automated Decision-Making in the NDIS and broader social services, comprising disability law experts, AI governance specialists, human rights scholars, and genuine DPO representation.

10. Require genuine co-design with Disabled People’s Organisations before any further eligibility, access, or support reforms are implemented, consistent with UNCRPD obligations.

11. Commission a full gendered economic analysis of the carer burden implications of proposed eligibility changes, including workforce exit, superannuation loss, and long-term financial impacts on women.

Submission 2773

Submission to the Senate Community Affairs Legislation Committee

9. Conduct a transparent fiscal analysis comparing NDIS reform “savings” with available alternative revenue measures, including reform of fossil fuel subsidies, before any further reductions to participant supports are implemented.

10. Closing Statement I have worked in disability services for 16 years. I have watched the small things that others miss: the impatience, the dismissiveness, the talking over, the lack of consent, the way humanity slowly erodes when people are seen as tasks instead of people.

Harm rarely begins with extreme cruelty. It begins with burnout left unsupported, accountability disappearing, and systems stretched beyond what is humane. History shows, repeatedly, that institutions are capable of profound human rights abuses when vulnerable people are no longer fully seen as human beings with equal dignity. One of the hardest things to carry is not the reality of disability itself, but the knowledge of how vulnerable disabled people can be within the very systems designed to protect them. Every parent in this position quietly wonders: will the world be gentle with them when I no longer can be?

A compassionate society is built in the answers to that question. In the systems we create. In the protections we prioritise. In whether dignity is preserved even when someone depends entirely on others for care.

Disabled people are not abstract policy discussions. They are not budget line items. They are deeply loved human beings, and families should never have to destroy themselves to access basic support and dignity.

Parent | NDIS Provider | Allied Health Professional

May 2026

References

Alon-Barkat, S & Busuioc, M 2023, ‘Human-AI interactions in public sector decision making: “Automation bias” and “selective adherence” to algorithmic advice’, Journal of Public Administration Research and Theory, vol. 33, no. 1, pp. 153–169.

Australian Federation of Disability Organisations (AFDO) 2021, The Economic Benefits of the National Disability Insurance Scheme and the Consequences of Government Cost Cutting.

Australia Institute 2024–25, Fossil Fuel Subsidies in Australia: Federal and State

Government.

Climate Council 2025, What is the diesel fuel rebate, and why is the government paying big miners to pollute?

European Data Protection Supervisor 2025, TechDispatch #2/2025: Human Oversight of

Automated Decision-Making.

Laux, J & Ruschemeier, H 2025, ‘Automation Bias in the AI Act: On the Legal Implications of Attempting to De-Bias Human Oversight of AI’, European Journal of Risk Regulation, vol. 16, no. 4, pp. 1519–1534.

Submission 2773

Submission to the Senate Community Affairs Legislation Committee

National Disability Insurance Agency (NDIA) 2025, Quarterly Report Q4 2024–25.

Per Capita 2024, Economic Contribution of the NDIS.