If 2025 was the year generative AI took off in organizations in every sector, it was also the year we saw increasing examples of the risks of AI mishaps. In 2026, AI is likely to move further into every part of our businesses – particularly with the development of digital agents. It pays, therefore, to look at what can go wrong and what compliance can do. 

Ruth Prickett graduated from Cambridge University with a BA hons in History and has specialized in business and finance journalism for the past 20 years. She was editor of Financial Management, the magazine...