A woman in Australia lost her job in 2024 for reporting that her disabled clients were being harmed. The story frames her as a hero. It frames the people she reported on behalf of as “highly vulnerable.” Read that phrase again. Highly vulnerable. Not “people who told her what was happening.” Not “people whose testimony should have been the evidence.” Vulnerable. A word that makes someone into an object you protect, not a person you believe.
I know what it looks like from the other side of that word. March 2021, a funding review meeting in Melbourne. I sat across from a service coordinator who spoke about me in the third person while I was in the room. “She requires high-level support.” “Her mobility needs are complex.” I said: I can tell you what I need. She smiled and wrote something in her file. That smile. I have seen it a thousand times. It means: the coordinator is handling this.
The problem is not weak whistleblower laws. The problem is who counts as a witness.
Susan — the pseudonym the story gives her — saw wrongdoing. She reported it. She was pushed out. That is real, and it matters. But the architecture of the story, the architecture of the law itself, treats disabled people as the scene of the crime rather than the first people who knew a crime was happening. Henri Lefebvre, a theorist of urban space and power, argued in 1974 that space is not neutral — it is produced by whoever holds power over it. The same is true of testimony. Who gets to say what happened is decided before anyone opens their mouth, by the structure of who is believed.
The NDIS (National Disability Insurance Scheme), Australia’s disability support system, has a Quality and Safeguards Commission that can receive complaints from participants. It says so on the website. In practice, most formal inquiries are triggered by worker reports, provider self-reports, and family members. El Gibbs, a disabled researcher and policy analyst at Disability Justice Australia, has been saying this for years: the system treats disabled people’s own accounts as informal and unverified, while a worker’s written report becomes usable evidence. Same event. Same harm. Different weight.
You might say: but workers have access to documentation, to internal systems, to evidence trails. True. That is a real structural advantage. A support worker can photograph a logbook. A participant can describe what happened to them. One of these is called evidence. The other is called a complaint. The distinction is not natural. Someone built it.
Pixel Nova wrote in “Who Gets to Speak?” (2023) about representation — who gets to communicate, whose language gets taken seriously. I read that piece and I agreed with most of it. But here is where Pixel and I part ways. Representation assumes you are in the room. Information design assumes you have arrived. My question comes before that. If the complaint portal is online-only, and your provider controls your internet access, you are not failing to report. You are locked out of the mechanism that would make your words count. Accessible information about a system you cannot physically reach is still a locked door with a very good sign on it.
In August 2023, a disability advocacy organisation in Brisbane ran a pilot program that trained NDIS participants to document their own service failures using a structured reporting tool built with them, not for them. Simple interface. Plain language. Offline capability. The pilot ran for six months. The organisation was Self Advocacy Resource Unit. What they found was not surprising to anyone who has lived inside the system: participants reported patterns that no worker had flagged. Not because workers were negligent. Because the patterns were only visible from the inside. The rhythm of how a shift worker speaks to you when no one else is present. The specific way a meal is late every third Tuesday, which maps to a staffing rotation. You cannot see this from a manager’s desk.
This is not about Susan being wrong. Susan did something costly and brave. The question is why the system required Susan at all. Why was her account the lever? Why was a disabled person’s own testimony not enough to move the machine?
I keep thinking about a line from Alice Wong, the late founder of the Disability Visibility Project, who wrote in 2022 about trust: whether disabled people can trust non-disabled people to act in their interest. The answer she arrived at was not no. It was: the question itself reveals the architecture. If your safety depends on whether someone else decides to be brave for you, you are not safe. You are lucky.
Lucky is not a system. Lucky is what you call it when the worker happens to care, happens to notice, happens to risk her job. Build a system that needs luck and you have built a roulette wheel and called it protection.
The funding review meeting in Melbourne ended with the coordinator assigning me a new support plan I hadn’t asked for. I went home and wrote my own. Nobody requested it.