The rise of AI is impossible to ignore. From education to defence and every sector in between artificial intelligence is rapidly changing how we access and interpret information.
In the world of tax and accounting, that change is already here.
Earlier this year, HMRC engaged two specialist contractors to develop advanced AI modelling tools. These systems are designed to flag discrepancies across vast data sources, spotting anomalies that may suggest tax has been underpaid, whether intentionally or not.
The impact? Significant. HMRC issued a record £16m in fines in the 2024/25 tax year to mid-sized UK businesses for corporation tax issues, many of which were prompted by data anomalies alone.
At Wilson Partners, we’re seeing this play out in real time.
In one case, a business sale was reported in the buyer’s accounts at £10m and in the seller’s tax return at £8m. Both figures were accurate. The difference reflected £2m in bank debt cleared during the transaction. A normal and explainable structure. But what AI flagged was a conflict, not context.
The result? A full-scale tax enquiry completely avoidable with the right interpretation.
We’re also seeing more clients turn to AI tools like Google or ChatGPT for answers. Sometimes they’re close. Other times, they miss critical nuances that can make all the difference.
There are two core issues at play:
- You only get the right answer if you ask the right question
- AI still makes mistakes
Tax legislation doesn’t always follow common sense. It’s complex, layered, and often shaped by case law and precedent, elements that are rarely captured in a simple AI output.
That’s not to say AI doesn’t have a role. But every AI-assisted output should be reviewed by an experienced advisor. Because in tax, small mistakes can have significant consequences.
Think of it like Googling a medical diagnosis. It’s useful but not something you’d rely on without speaking to a professional.
The same principle applies here.
AI is powerful, but it’s most effective when paired with sound judgement, sector knowledge and experience. In our world, it’s not about man versus machine. It’s about using the best of both to deliver better outcomes for our clients.
And in a world where almost everything leaves a digital footprint, it’s never been more important to get it right the first time.
