AI and the changing fraud landscape

Artificial intelligence (AI) is transforming the way fraud is committed and detected. A recent Home Office briefing highlights how criminals are using AI to make scams more convincing, more scalable and harder to identify.

The Home Office recorded a 395% increase in reports involving the suspected use of AI during 2025-26, with associated losses rising from £1.2m in 2024-25 to £9.6m in 2025-26. While AI-enabled fraud still represents a relatively small proportion of overall fraud reports, the trend points to a rapidly evolving threat.  

AI is being used to create fake identities, generate persuasive communications, produce fraudulent documents and websites, and support increasingly sophisticated impersonation attempts. Technology is also allowing criminals to automate activities that previously required significant time and expertise.  

For accountancy practices, these developments reinforce the importance of robust client due diligence, effective identity verification and ongoing staff awareness. As AI-generated content becomes more convincing, practices should be alert to the risk of increasingly credible fraud attempts targeting both clients and employees.  

The response is evolving too. Government, law enforcement and industry are investing in AI-enabled tools to identify, disrupt and prevent fraud, with initiatives including the Fraud Strategy 2026-29 and the new Online Crime Centre.

How to protect your practice 

AI doesn’t change the fundamentals of fraud prevention, but it does raise the stakes.  

It’s making fraud more convincing, but practices that combine strong controls, informed staff and effective use of technology will be better placed to protect their clients, their reputation and the public interest.   

You should review your client onboarding and verification processes, help staff understand emerging AI-enabled fraud risks, and encourage them to challenge unusual requests or behaviour. As fraudsters adopt increasingly sophisticated technology, professional scepticism remains one of the most effective safeguards.  
  
 Contact the Practice Support team using the technical helpdesk for more information and support.