A Million-Dollar NDIS Plan: Who Is Being Helped, and Who Is Being Paid?
An adult with severe autism reportedly has more than $1 million a year allocated for NDIS support. A former worker says he and a colleague once helped him explore beaches, develop new skills and experience very few serious incidents during their support. Their question is simple: why should a more expensive service model be accepted without proving it produces a better life?
Australia spends around $50 billion a year on the National Disability Insurance Scheme. That is serious public money, and Australians are entitled to ask where it goes. But there is a difference between cutting the NDIS and cutting the waste that prevents it from doing its job.
One reduces support for people with disability. The other demands that funded services deliver something worthwhile. It also asks why a registration certificate can carry more weight in support decisions than years of demonstrated competence with the individual.
Consider the experience of a nonspeaking autistic adult, called Eli here to protect his identity. A former independent support worker says Eli's annual NDIS allocation is well over $1 million. Such funding may be justified when a person needs continuous intensive care. But the worker believes the package could have been significantly more effective—and potentially less costly—if more weight had been given to the outcomes achieved by two long-term supporters.
He was not just surviving. He was living.
Eli had a history of severe distress. He sometimes shouted, struck out or struggled when his expectations changed. Those incidents could be frightening and required sensible safety arrangements.
Yet two familiar support workers found ways to help him enjoy a far broader life. They learned how he communicated. They gave him time to process change. They took him into the community, encouraged new experiences and helped develop practical skills. The former worker describes adventurous outdoor activities, increasing independence and a level of trust that enabled experiences other workers considered too difficult.
Serious incidents during their own support were reportedly rare. On many activities, one experienced worker was able to support Eli without additional people shadowing them.
His disability had not gone away. What changed was the way he was supported.
That should have made everyone curious. If a person can be settled, safe and engaged with one team but repeatedly distressed in another arrangement, the difference is information. It might tell us something about communication, training, environment, expectations or staffing. It might also reveal risks not visible during the successful shifts. Either way, the question deserves investigation—not just another funding justification.
A call for help becomes an incident
Eli had developed a short sound that meant the name of a trusted worker. When he struggled to communicate with others, he sometimes called for that person.
According to the former worker, staff understood whom he was asking for but treated the calling as a behaviour of concern.
Here lies the difference between counting behaviour and understanding it.
A nonspeaking adult calling for somebody who understands him may be trying to solve a problem. If that request is ignored and frustration escalates, the resulting incident may then be used to justify more intervention, more supervision and more cost.
Not every episode of aggression has a single explanation, and no supporter should ignore risks. But a service paid to understand and assist a person ought to be able to tell the difference between communication that needs responding to and behaviour that simply needs suppressing.
Registered on paper. But does anyone know the person?
There is another uncomfortable question in Eli's story. His two most familiar supporters were independent workers who had spent years learning his communication, moods, routines, preferences and signs of distress. They were not necessarily the workers backed by the largest organisation or the most impressive compliance system. They were the people who knew him.
The former worker says staff turnover in Eli's provider-managed supported living arrangement was high and that some incoming staff lacked the specific preparation needed to understand him. He describes a revolving door of unfamiliar people entering the life of someone for whom familiarity and predictable communication were essential. The providers concerned would need to respond to that account; worker induction records, rosters and incident reports could establish how often staff changed and what preparation was supplied.
For a nonspeaking autistic person, that is not an administrative inconvenience. Communication can depend on a supporter recognising the meaning of a sound, a repeated phrase, a gesture or a change in expression. It can take months or years to build that understanding. When a worker leaves, the person's communication system can lose a vital partner. When replacements arrive without enough knowledge, the person may have to start over again and again.
Why should a registration certificate carry more practical weight than the demonstrated ability to understand the person it is supposed to protect?
Under the NDIS, it is the provider—the business or organisation—that registers. Registration brings auditing, screening, incident-management requirements and oversight. Those protections are important, especially in services involving high support needs or restrictive practices. But registration does not mean that every employee has spent time learning an individual participant's communication or knows how to respond when that person becomes distressed.
In 2026 the government expanded mandatory registration to Supported Independent Living (SIL) providers. From 1 July, new SIL providers required registration; existing unregistered providers had a transition period and had to apply by 1 October to keep delivering SIL. The government says the reform is intended to strengthen safeguarding and accountability. The Commission's guidance makes clear that provider-managed SIL is subject to these requirements.
But a participant who chooses, directs and rosters their own support workers is not automatically receiving provider-managed SIL under that definition. The NDIA explicitly recognises that distinction. Many self-managed and plan-managed participants can still use independent, unregistered workers for supports where registration is not mandatory. Crucially, the NDIS Code of Conduct applies to registered and unregistered workers alike.
The problem arises when registration is treated as a shorthand for competence, or when administrative and funding arrangements leave people with too little practical choice. A person may be moved away from an experienced independent supporter and toward a registered organisation whose workers have passed general screening but do not yet know that person at all. The result may satisfy the paperwork while undermining the very continuity and understanding on which safe support depends.
Indeed, the Commission's new SIL practice standards expect workers to understand each participant's specific communication style, preferences and needs, with consistent practice across workers and shifts. That is an excellent requirement. The question is whether it is happening at the front door, not simply written in the organisation's policy folder.
No one should have to choose between proper safeguards and the people who genuinely understand them. Registration should make good support accountable, not make demonstrated good support disposable.
In Eli's case, the former worker's departure predated the latest SIL registration rules, so those reforms cannot be assumed to have caused it. But the case exposes the risk any expanded registration regime must confront: when the rules determine which organisations may provide support, they must not erase the participant's right to continuity, genuine choice and relationships that work.
The hidden cost of starting again
High turnover can make even an apparently compliant service expensive in ways invoices do not reveal. New workers require induction. Familiar routines are disrupted. Communication breaks down. Distress can increase. More staff may be rostered to manage risks that better continuity and individual knowledge might have prevented.
Those costs should not be assumed in every case, and staffing decisions must always reflect assessed safety needs. But they are measurable. A serious NDIS value-for-money review should compare the cost of turnover and repeated escalation with the outcomes achieved by stable teams. It should ask not only how many workers attended, but whether the participant became safer, more understood and more able to live the life they wanted.
For someone like Eli, the difference between an unfamiliar roster and two trusted communication partners may be far more than emotional comfort. It may be the difference between an expensive cycle of crises and a life of participation.
When funding rewards the wrong evidence
The former worker later lost his position. He says opportunities and familiar communication strategies were reduced after support arrangements changed. He also alleges that documentation showing Eli's capabilities was discouraged or altered because it might weaken the case for ongoing funding.
A dossier supplied to the safeguarding regulator identifies historical messages raising questions about whether periods of improvement were fully recorded when funding was under review. Those original records and the final support reports deserve independent comparison.
The wider problem is not difficult to understand.
Funding is often justified through evidence of need. That is sensible. But when reporting concentrates on the person's worst days, there is a risk that progress starts to look financially inconvenient. A person who learns to regulate distress, communicate more clearly and enjoy ordinary community life may appear to need less support on paper, even though effective support is precisely what made the improvement possible.
So what gets rewarded: the skilled work that prevents a crisis, or the expensive response after the crisis happens?
A system that makes the management of recurring incidents easier to fund than prevention creates an opportunity for waste. If poor outcomes can be continually used to justify larger packages without meaningful scrutiny, the safeguards are inadequate. If anyone deliberately distorts records or claims services that are unnecessary or undelivered, that becomes a matter for investigation and enforcement—not a reason to punish every participant.
The auditors have already found the weaknesses
This is not a debate based entirely on one worker's account. Australia's Auditor-General found in June 2025 that the NDIA's management of claim compliance was only partly effective. Until 2024, the scheme lacked basic preventative controls such as strong identity checks and pre-payment validation. The audit reported that in 2023 the NDIA estimated fraud, non-compliance and payment errors at 6–10 per cent of outlays.
The figure covers different forms of financial leakage; it is not a finding that those billions were stolen by support workers. But it is a compelling reason for taxpayers to demand better controls.
The 2023 independent NDIS Review identified another problem. Participants said planning and budget decisions focused on what they could not do rather than what they could. That mindset may encourage assessments of incapacity while neglecting evidence about which supports produce genuine improvements.
What kind of system asks people to document how badly life is going, but struggles to recognise the value of the support that makes life better?
The tobacco lesson: follow the incentive
Australians have seen a different policy produce a similar warning about unintended incentives.
Tobacco excise was intended to discourage smoking. Yet as legal cigarettes became much more expensive than illicit ones, a lucrative illegal market expanded. The Australian Bureau of Statistics has documented the sharp shift towards cheaper illicit tobacco. The tax was not created to fund organised crime; the price gap nevertheless created an opportunity that criminals exploited.
The NDIS is not a tobacco market, and its funding problems are different. But the lesson carries across: when incentives are poorly aligned with a policy's purpose, money can flow towards the wrong outcomes.
In disability support, an incentive to maximise billable hours rather than useful outcomes can encourage overservicing. An incentive to present somebody at their most incapable can distort the evidence needed to make fair decisions. Weak compliance systems make deliberate fraud easier. None of these failures is inevitable; they are reasons to design the rules better.
The million-dollar question no one should dodge
In Eli's case, the former worker believes two experienced supporters, backed by appropriate specialist input, relief staffing and independent safeguards, could deliver substantially better results with fewer unnecessary layers of supervision.
That proposition needs a genuine assessment. Two people cannot cover continuous care by themselves, and reducing staff without considering safety would be irresponsible. But neither should an existing staffing model receive a blank cheque simply because it is established.
Compare the actual costs. Compare staffing hours. Compare incident rates, restraints, activity participation and communication outcomes across the different arrangements. Look at what Eli could do before and what he can do now. Identify what has changed and why. Then decide what level of support is genuinely necessary.
The point is not that a million-dollar plan must be cut. It is that a million-dollar plan deserves a million-dollar level of accountability.
Better management is not an attack on disability
The most important distinction in this debate is between people who need support and systems that may be failing to deliver it.
Disabled Australians should not lose essential services because governments have tolerated weak oversight or because a provider has exploited a loophole. Likewise, responsible workers should not be blamed for the actions of the minority who overcharge, falsify records or fail the people they are paid to assist.
Public accountability can be practical: independently check claims against delivered services; record successful prevention alongside incidents; audit high-cost arrangements against actual outcomes; measure turnover and participant-specific staff competence; preserve established communication partners wherever safe and appropriate; investigate credible concerns regardless of whether they come from a registered company or an independent worker; and ensure an improving participant does not automatically lose the supports responsible for that improvement.
The former worker in Eli's case has lodged a detailed complaint with the NDIS Quality and Safeguards Commission. He has sought an account of its progress and outcome. His concerns, and the conduct of the services involved, warrant a fair and independent examination.
The stakes extend beyond an individual dispute. A poorly managed NDIS can be both extraordinarily expensive and miserably ineffective for the person at its centre.
Australians should not have to choose between protecting vulnerable people and protecting public money. The NDIS was supposed to do both.
Cutting genuine need is easy to announce and devastating to experience. Eliminating unnecessary spending, misleading reporting and poor-quality support is harder—but that is where sound management belongs.
When a person's life gets smaller while the bill gets bigger, the first question should not be how much more money the system needs.
It should be: who is being helped, who is being paid, and what is the money achieving?
Editorial note and invitation: This is a journalist-style public-interest commentary based on a former support worker's account and supporting correspondence, with the participant and identifying details altered. It is not presented as a completed independent newsroom investigation. The reported annual plan exceeding $1 million, proposed savings and allegations about individual conduct have not been independently established. Journalists interested in examining original records and speaking with the people involved are invited to request a confidential briefing at thesmilingmonad@gmail.com. Any investigation should protect the participant's privacy and seek responses from relevant parties.
The Smiling Monad · 9 November 2026
