The Financial Conduct Authority (FCA) intends to integrate AI into its monitoring and disciplinary operations, operating as an “agentic regulator” while holding off on creating dedicated rules for the sector, reported the Financial Times.
Jessica Rusu, the watchdog’s chief data, information and intelligence officer, will present the organisation’s position on frontier AI models during a speech scheduled for next week.
Go deeper with GlobalData
Access deeper industry intelligence
Experience unmatched clarity with a single platform that combines unique data, AI, and human expertise.
The planned address arrives as supervisory authorities assess both operational applications and institutional risks linked to advanced systems.
Rusu’s presentation will outline the regulator’s response to proposals put forward in July by Sheldon Mills, formerly a senior executive at the FCA.
The Mills Review advised authorities to expand their statutory remit and determine whether large language models including ChatGPT, Claude, and Gemini warrant direct oversight under existing financial rules.
Supporting survey data in the review showed that 20% of adults in the UK would permit AI systems to manage financial choices, including personal borrowing and savings allocation.
FCA chief executive Nikhil Rathi remains doubtful regarding the necessity of statutory additions for AI.
During testimony before the Treasury select committee this week, Rathi stated that although his perspective “has been evolving”, he requires proof that current regulatory instruments are inadequate before seeking “lots more powers until we have really demonstrated that we have used our powers to the fullest extent”.
Addressing frequent inquiries on whether the watchdog should police AI systems, Rathi detailed the current legal boundaries.
The review separately advised the authority to “build and adopt an AI-enabled agentic supervisory model” to maintain oversight as market operations shift towards autonomous software requiring diminished human intervention.
At the committee session, Rusu noted that AI tools assist staff in processing and condensing “unstructured text” submitted by firms across records, electronic mail, and formal interviews. The regulator has conducted on-site evaluations at financial institutions to observe active deployments of the technology.
