The ATO’s latest Top 100 findings report, released on 17 September, opens with good news. 82% of the Top 100 hold a high or medium assurance rating for income tax. For GST, it’s 98%. Ten years after justified trust began, the program works.
Read past the headline and the message is sharper. The ATO is no longer asking whether your tax governance exists. It is asking you to prove that it runs.
THE BAR HAS MOVED FROM DESIGN TO OPERATION
The ATO rates tax governance in stages. Stage 2 means your control framework is designed effectively. Stage 3 means it also operates effectively in practice, and for that the ATO wants evidence “that a taxpayer has independently tested the operation of its framework and it is found to be operating effectively.”
This year 49% of the Top 100 reached Stage 3, up from 41% last year and one in five in 2023. 76% now hold at least a well-designed framework. Design no longer separates the leaders. Operation does.
GST is further behind. Only 16% of the top 100 GST reporters reviewed have reached Stage 3.
A design is a document. Operation is evidence, produced every time a control runs.
WHERE THE EVIDENCE RUNS OUT
The report is specific about what holds taxpayers back. The areas most commonly lacking documentation in 2026 were:
- Board reporting. Who reports tax risks to the board, what they report, how, and how often.
- Book-to-tax differences. Procedures explaining significant differences between accounting and tax outcomes.
- Controls testing. A periodic internal controls testing program.
On testing, the ATO names exactly what stopped taxpayers reaching Stage 2: no detailed testing plan covering every control, no documented testing frequency, and no evidence the board or a delegated committee endorsed the program.
None of these is a disagreement about tax law. Every one is a gap in evidence. That’s not a technical problem. It’s a systems problem.
Think about where that evidence lives in most tax functions. The book-to-tax walk sits in a workbook rebuilt every year. The review happened, but the sign-off is an email. The board paper was assembled by hand from five files. Every control may have run. Proving it means rebuilding the trail after the fact, and an independent tester can only test what they can find.
ASSURANCE IS NOT SET-AND-FORGET
That’s the ATO’s phrase, not ours. A small number of Top 100 taxpayers were downgraded this year. High assurance grew from 55 taxpayers to 57, but low assurance rose too, to 14 taxpayers, or 15%.
The engagement model has moved too. Over 95% of the Top 100 now have real-time justified trust reviews underway, and 80% have no past-year reviews outstanding. Most of the monitoring reviews the ATO completed “identified one or more new significant transactions that required review.” Its request is plain: make sure “evidence supporting the tax treatment of significant transactions is readily available and provided to us in a timely manner.”
Real time changes the economics of evidence. If the ATO is looking at this year while it happens, the evidence has to exist while the work happens. Assembled at review time, it becomes a scramble every year.
EVIDENCE AS A BY-PRODUCT OF THE WORK
This is the problem we built TaxTime to solve, so it’s worth being specific about what a system changes.
The reconciliation becomes the procedure. In a tax ledger, every difference between accounting and tax is recorded once, against the account, the rule and the workpaper that produced it. The ATO says the detailed book-to-tax reconciliations it now receives “allow us to obtain assurance over the key adjustments.” When that reconciliation is a live calculation rather than an annual spreadsheet, explaining a significant difference stops being a document you write. It becomes a view you open.
Controls leave their own trail. Role-based access, maker-checker workflow and an audit log that can’t be edited mean every review and approval is recorded: who, what and when. Your testing program samples from that record instead of chasing inboxes.
Board reporting comes from the ledger that computed the numbers. Positions and evidence travel with the pack, so each one is a record of what the board was told and when.
Significant transactions are documented as they happen. The workpaper, the position and the supporting evidence sit together, dated, ready when the ATO asks.
The principle is simple. The output of running the process is the evidence set.
THE NEXT TEN YEARS
The first decade of justified trust was about earning the ATO’s trust. The next is about keeping it, every year, in real time, with evidence that stands up to independent testing.
A controller asks: how do we get through the next review? A CFO asks: how do we make every review routine?
It’s also the foundation for AI in the tax function. An agent’s work is only as reviewable as the system it runs in.
If a refresh review is on your calendar, start with the three gaps the ATO named. For each one, ask a simple question: where does the evidence live today?
Off the Tax Treadmill. On to Impact.
