The NDIS Says It's Fighting Fraud. Its Own Numbers Point at Providers, Not Us.

On 15 June, the NDIA published a short update. From 18 June, they said, they'd be doing extra checks on older claims. First, anything submitted more than 12 months after a support was delivered. Over the coming months, that window shrinks to six. Claims caught in the net can be held for up to 28 days while they're reviewed.

Read on its own, it sounds sensible. Nobody wants NDIS money going to fraud. I don't either. My family relies on this scheme, and every dollar that gets siphoned off by someone gaming the system is a dollar that isn't there for a kid who actually needs it. But the more I sat with this announcement, the more it started to feel like something else entirely: a policy that punishes the people who are already doing the hard, unpaid, invisible work of self-managing a plan, while the money genuinely being taken sits somewhere else.

Who actually gets caught by a timing rule

Think about what it actually takes to submit a claim within six months, every time, for every support. You need to receive the invoice. You need to check it's correct. You need to log into the portal, upload the evidence, match it to the right support category, and file it before you forget. Now do that while you're also coordinating three or four different providers, managing a diagnosis, dealing with a school that doesn't understand your child, or simply surviving a period where life got harder than usual.

A timing rule doesn't measure whether a claim is honest. It measures whether a family had the bandwidth to file paperwork fast enough. Those are not the same thing, and the NDIA's own self-management guide makes this obvious: self-managed participants are explicitly told they can pay out of pocket first and claim reimbursement later. That's the system working as designed. Now the same system is treating the natural lag in that process as a red flag.

Meanwhile, a provider who wants to pad an invoice, bill for something that never happened, or quietly inflate a support need has every incentive to claim quickly, cleanly, and within the window. Speed isn't evidence of honesty. It just means you weren't caught by this particular filter.

The NDIA's own numbers say the fraud is on the provider side

This is the part that made me angry enough to write this. The NDIA doesn't just claim fraud is a problem, it publishes numbers about where its own enforcement has actually landed. Since the Fraud Fusion Taskforce was established in November 2022, it has disrupted more than 2,500 providers who submitted incorrect or non-compliant claims, or who showed other significant risk indicators. Close to 200 individuals and providers have been banned by the NDIS Quality and Safeguards Commission. Eighty-eight banned providers alone had claimed $28 million in the twelve months before they were banned.

The taskforce reports $2.2 billion diverted away from problematic providers, and more than $880 million saved by blocking non-compliant payments before they went out the door. Every one of those figures points the same direction: when the NDIA's own integrity machine goes looking for wrongdoing and actually finds it, it is overwhelmingly finding it on the provider side of the ledger.

So why is the newest layer of scrutiny a blanket timing check that catches every self-managed participant, plan manager and provider by how late they claimed, rather than a targeted response built on the risk profile the NDIA's own data already shows it?

The $3.7 billion figure everyone quotes, and what it's actually made of

You'll have seen the headline number: roughly $3.7 billion, or 8.3% of payments, described as affected by 'integrity leakage' in the NDIA's own estimate for the prior financial year. It sounds enormous, and it gets used to justify exactly this kind of crackdown. But read the fine print and that figure isn't a fraud number. It's a blend of genuine fraud, honest mistakes, and poor record-keeping, all lumped into one frightening total.

That matters enormously, because a family who submits a claim seven months late, or forgets to attach the right invoice format, gets counted in the same bucket as a criminal syndicate setting up a fake provider business. One of those is a paperwork error. The other is theft. Folding them together lets the government point at a huge number and imply it's all fraud, while the policy response, a blanket timing filter, is only capable of catching the paperwork errors, not the theft.

Every provider charges an NDIS price. That's not an accident, it's the system.

Here's the part nobody in government seems willing to say out loud. Once a funding label exists, the market prices to the label. The independent NDIS Review analysed real billing records in 2023 and found that 83% of all price-capped transactions happen at or near the maximum allowed rate. Not the average rate. The ceiling. For the vast majority of NDIS-funded support, the price cap isn't a safeguard against overcharging, it's the price everyone quotes.

Look at what that ceiling is actually built from. The 2025-26 price limit for standard weekday support work is $70.23 an hour. The worker delivering that support is paid $34.58 an hour under the SCHADS Award. That's a $35.65 gap, supposedly covering superannuation, insurance, administration, compliance and margin, on every single hour, for every single participant, almost every single time. The NDIA's own 2024/25 benchmarking found therapy rates exceeding comparable market prices by up to 68%.

This isn't a handful of bad actors rorting the system. It's the system. Participants can't shop around the way you would for any other service, because support needs are specific and options are limited by geography and disability type, and because the participant bears no direct cost at the ceiling price, there's no competitive pressure pushing prices down. Once you set up a business model once, around a price that's guaranteed to be paid, it becomes a work process. Charge the cap, every time, to everyone. That's not fraud in the criminal sense. But it is money quietly leaving disabled people's plans faster than it needs to, at scale, every single day, and it never once gets a compliance blitz, a 28-day hold, or a headline.

Participants are usually the ones being scammed, not the ones scamming

The picture gets worse when you look at who fraud actually happens to. Reporting on how NDIS fraud really operates describes participants being intimidated or threatened into letting a provider use their plan, criminals impersonating providers or support coordinators to redirect funds, and fake businesses set up purely to submit false claims against real people's plans. The participants caught in this aren't the perpetrators. They're described as frightened to speak out, worried they'll be blamed, or scared they'll be asked to repay money they never controlled in the first place.

That is who this scheme should be protecting first. Instead, the newest enforcement mechanism is a timing filter that treats every self-managed participant, plan manager and provider the same way, regardless of whether they've ever done anything wrong, based purely on how many months have passed.

What we actually pay for that ordinary providers never see

I want to bring this back to my own family, because the numbers only mean something when they're attached to a real life. We get NDIS funding for our son. None of it is spent on luxuries. It goes to Occupational Therapy, Speech Pathology, Psychology, social workers helping him build the skills to navigate a world that wasn't built for him. We pay out of our own pocket for plenty of it too, exactly the way the NDIA's own guide tells self-managed families to do. Every cent is accounted for. Every receipt is kept. And every provider we've ever dealt with quotes at, or close to, the NDIS price cap, not the price they'd charge someone paying privately for the same hour of their time.

So when I read that the scheme is tightening the screws on timing, on the exact administrative lag built into the way self-managed families are told to operate, while its own data shows the real dollars are walking out the door through provider pricing and provider fraud, it doesn't read as fraud prevention to me. It reads as the easiest people to police being policed first, because we're the ones who show up, keep records, and answer the letter when it arrives. The syndicates setting up fake provider businesses are harder to find. Families doing the right thing are not.

What should actually happen

I don't think most people working inside the NDIA are trying to punish families like mine. I think a genuinely hard problem, real fraud, in a scheme that has to serve hundreds of thousands of people, has produced a blunt policy that's easiest to apply to the people least likely to fight back. But easiest isn't the same as right. If the government wants to protect the integrity of this scheme, the evidence it has already published tells it exactly where to look. It's not families filing a claim seven months late. It's the price on the invoice, and who set it there.

References