One day, the announcement will simply say, "We now do a full set of accounts." No consultation period, no transition plan for the profession, no runway. The question is not whether production automation reaches statutory accounts work. It is whether your firm meets that moment already running, or reacting to it.
On 25 August, Xero's £120-a-year accounts and corporation tax product went straight to business owners. Intuit put QuickBooks inside ChatGPT last year. Two moves is a pattern, and the pattern doesn't stop. The government's free CATO service closed on 31 March; the state stopped doing it free, and the platforms are selling it to your clients. AI has removed the production excuse. The answer is not to race a platform to the bottom: it is a properly reviewed, properly defensible file at a cost that makes the comparison silly, plus the time to do the work a platform can't.
The timing case deserves better than a landing page summary, so we have not tried to compress it. The Last Manual Year sets out the platform pattern, what happened to the legal profession's timetable, and why waiting to see is the one strategy with no upside. It names no products, including this one; it is an argument about the market, not an advert.
If the paper's argument holds, the next step is not believing ours. It is testing it: run the numbers on the calculator with your own figures, or hand the engine ten finished jobs and see what a cold review finds.