The AI regulation challenge

Use of artificial intelligence is transforming many operations within financial services. Yet, while AI delivers game-changing benefits, its potential harms are the subject of increasing regulation. Minesh Tanna outlines the onerous regime proposed by the European Union and warns that compliance with AI laws is not straightforward.

Minesh Tanna is a partner and global AI lead at law firm Simmons & Simmons. He is also chair of the Society for Computers and Law (SCL) AI Group. Contact him on minesh.tanna@simmons-simmons.com.

Generative AI (in particular, ChatGPT and Bard) has recently been attracting growing attention. This is unsurprising: these are powerful examples of sophisticated AI systems (specifically, large language models or 'LLMs') which can produce informative and creative content with minimal human instruction. However, these LLMs, and AI more generally, are not infallible. AI systems - which, in advanced form, are complex, opaque and unpredictable - are capable of causing harm if not properly developed or used.