TaxTime Blog

Two heads of tax told me their teams aren’t adopting AI as fast as they’d like. The usual reading is that tax people are careful with new tools. The better reading is that their AI has had nothing to reach. Adoption pace is an input. A strong tax function is measured on results: adjustments explained, significant transactions documented before lodgement, governance evidenced in real time, and the share of the week spent in the business before decisions are made.

Most Top 1,000 taxpayers the ATO has assured sit at medium. The ATO’s 2026 findings set out what it takes to reach high: governance that is designed and operating, and evidence an independent tester can find.

The ATO’s latest Top 100 findings report opens with good news: 82% of the Top 100 hold a high or medium assurance rating for income tax. 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 market read the last decade as tax moving out of the building. The evidence moved the other way. Justified Trust changed what the regulator asks for. Advisor dependency became a risk boards price. Neither was a change in cost. This article is about the distinction that follows: use the firms for judgement, and keep the ledger, the reasoning and the evidence in the building. You should be able to change advisors without losing your memory.

Most "AI in tax" answers skip the operating layer. They put a chatbot in front of a treadmill with no constraint identified, no inspection point, no pull signal, no feedback loop. Ninety days into most pilots, the team goes back to the spreadsheets. The model was not the problem. The function it was dropped into was not ready to receive it. There is a third path: build the harness first. Five operating laws, drawn from Goldratt, Grove, Toyota and Deming, that make a tax function able to absorb an agent.

Most conversations about AI in tax start with the wrong question. They ask which model is best, or whether a chatbot can read a tax act. Those are interesting questions. They are not the question. The real question is whether a tax leader can ask an agent something useful about the numbers and trust the answer. For almost every Australian tax function today, the answer is no, and the reason is the architecture, not the model.

A Head of Tax sat through three vendor demos last month. Every demo started with the word AI. Every demo ended with a chat window. The product was not changing. Only the label was. This article gives the buyer a vocabulary. Four levels of AI in tax. One question that cuts through the marketing. A way to walk out of a vendor demo knowing exactly what you were just shown.

Every tax team has a Sarah. Key person dependency is not a personnel issue, it is a structural one. This article looks at how operating models and technology choices either concentrate institutional knowledge in one person’s head or convert it into something the organisation owns.

Most build-vs-buy conversations miss the actual question. Because ‘build’ isn’t one thing. And what you’re really buying isn’t software. It’s the evidence that your tax position was produced by a controlled process.

The prevailing assumption in corporate tax transformation is that automation is always the goal. But not every workpaper is a fixed asset schedule. And the gap between this could be automated and this should be automated right now is where most transformation projects go wrong.

A governance framework that lives in a drawer isn’t governing anything. This article explores the gap between having a tax governance framework and actually living it, and how embedded controls in purpose-built software bridge the divide.

You can’t build trust on spreadsheets any more than you can build a house on sand. Most organisations have a gap between the governance they claim and the tools they actually use. Purpose-built tax software closes that gap across all four Justified Trust pillars.

The ATO isn't trying to catch you out. They're trying to decide whether they need to. Justified Trust isn't about catching mistakes. It's about the ATO determining whether it can trust your tax outcomes without needing to audit every line.

The ATO doesn’t want to know your answer. They want to know how you got there. When a Justified Trust review arrives, the difference between a calm response and panic comes down to one thing: an audit trail.

Your general ledger tells you what happened. Your tax ledger tells you what it means. Most organisations translate GL data into tax data manually every year: in spreadsheets, in someone's head, with no audit trail. A tax ledger captures this translation once, making it documented, traceable, and repeatable.

Free tools are the most expensive tools you own. Excel is free, but the gap between what it is and what you need it to be has a cost. A large one.

This story repeats itself across the country, in companies of every size. Not because the software is bad, but because the implementation ignored the human beings who were supposed to use it.

The spreadsheet you use to justify replacing your spreadsheets had better be convincing. Here's what actually matters when measuring ROI on tax software, and what's a distraction.

You already know you need it. Here’s how to prove it to everyone else. Most Heads of Tax have a broad mandate to drive digital change, but having the mandate and having the tools to act on it are different things.

The organisations that succeed at implementing tax software share a common trait: they treat the implementation as a transformation project, not a software installation. Here's what a successful implementation actually looks like.