The debate
Accounting is being reshaped by AI. The real question is not the one you think.
The question is not “does your accountant use AI?” Everyone will claim they do. The question is who owns the intelligence your work generates — and who owns the judgement when it matters.
The changes already underway
Four shifts — movements, not predictions
From hearing from your accountant at lodgement, to being told what matters before you ask.
From paying a monthly fee to a vendor who keeps the intelligence, to owning the system your spend generates.
From a once-a-year human glance, to a system that reads, re-checks and improves on a schedule.
From trusting a confident PDF, to seeing where every number came from — and what could not be determined.
It gets sharper on its own
Sharper than last quarter — and nobody had to tell it to be
We do not publish a vanity percentage. We publish the mechanism, and the honest fact that it runs on a schedule whether or not anyone is watching. A once-a-year human review cannot compound like this.
The mechanism. Every engagement adds structured signal. The engines re-weight against it, so the next answer starts further ahead than the last. Compounding, not catch-up.
Reads real demand, daily
The keyword and content engine reads what the market is actually searching for and publishes to it — the library grows itself, not once a year.
Re-checks its own work, on a schedule
The SEO, AEO and GEO engine re-audits and rewrites metadata across the whole network on a schedule — errors are caught by the system, not a spring clean.
Shares what it learns, across the network
Mission Control passes a lesson learned in one part of the practice to every other part — learned once, applied everywhere.
Direction, not a vanity number
The honest version: this is cadence and mechanism, not a headline number. The point is that it improves without being told — and that the compounding belongs to you, not a vendor renting it back.
The questions worth asking
What to ask any firm that says it uses AI
How do I know if my accountant is keeping up with AI?
A fair test is not whether they mention AI, but what they can show you. Can they tell you something before you ask, rather than only at lodgement? Can they show the working behind a number, not just the answer? And do they own the intelligence your work generates, or rent it from a vendor who keeps it? If the honest answer to those is “no”, the tools may be new but the way of working is not.
Should my accountant show me the working, or just the answer?
Both. An answer you cannot interrogate is an answer you are trusting on faith. A professional standard is to trace where each number came from and, just as importantly, to state plainly what could not be determined and hand that to a human for judgement. Publishing the limits is more credible than a confident PDF that hides them.
Who owns the automation and intelligence my accounting spend is generating?
This is the question most owners never ask. If your books are coded by an overseas platform on a monthly fee, the automation — and the intelligence it builds about your business — belongs to that vendor, not to you. The alternative is a practice that builds its own platform and lets you own the compounding. Ask who keeps the intelligence when you leave.
What does “audit-ready” actually mean for an NDIS provider?
Audit-ready means the evidence is in order today — worker screening current, claims within price caps, incidents logged and mapped to the Commission’s obligations — not assembled in a panic when the auditor calls. Audit-hopeful is the opposite: compliant in belief, untested in evidence. The difference only shows up under review, which is exactly the wrong time to find out.
When should founders paper a shareholders’ agreement?
Before it matters, not after. While co-founders are aligned, the cost of agreeing valuation method, buy-sell terms, leaver provisions and deadlock resolution is low. Once someone wants out — or there is a death or a dispute — the same conversation is expensive and adversarial. The cheapest insurance a growing business can buy is the paperwork it hopes never to need.
Is “AI accounting” just a chatbot bolted onto a website?
Often, yes — a wrapper over a generic model that answers questions but does no work and owns no judgement. That is one rung of a ladder. Further up sit systems that actually do the volume — the reconciliation, the first draft, the compliance check — under a principal who reads the detail and owns the call. The label “AI” tells you nothing; what the system does, and who owns it, tells you everything.