AI is dominating the FCA's newsfeed this week.

Today, the FCA has published a speech entitled “AI for growth – how the FCA can help” by Jessica Rusu, FCA chief data, information and intelligence officer.

It comes as no surprise given her role that Rasu has “a pro-technology approach to making the FCA a smarter regulator”, one of the key pillars of the FCA's new Strategy 2025-2030.  So, what does this mean in practice?

How is the FCA using AI?

• Predictive AI in the Supervision Hub: To assist agents with real-time knowledge. Presumably, this refers to the use of AI to locate pertinent information within the large volumes of data it holds on the firms' it regulates.
• AI Voice Bot, i.e. an AI chatbot: Uses AI to simulate human-like conversations to help direct consumers to the appropriate agency, be that the FCA or rather the FOS (Financial Ombudsman Services) or FSCS (Financial Services Compensation Scheme).
• Experimental usage of Large Language Models (LLMs):  Applying LLMs to process and interpret unstructured data, identify patterns, and enhance decision-making and streamline the FCA's Authorisations and Supervisory processes. 
• Advanced AI-analytics: Applying AI-enabled analytics specifically to voluminous trading data to identify potential misconduct requiring further investigation.
• FCA's AI Lab: The AI Lab fosters collaboration between the FCA, financial firms, and tech innovators to explore and implement AI solutions that enhance regulatory processes and ensure compliance.

What is in the pipeline?

Today, the FCA has announced the launch of AI Live Testing, as part of the AI Lab.  Through AI Live Testing:

“firms will be able to live-test AI models to build confidence in their performance, while also receiving regulatory support and comfort to support safe and responsible deployment”.

The details of the FCA's proposals are set-out in their Engagement Paper.  Responses are due by 10 June 2025.

What about AI policy?

In short, the FCA plans to use its existing regulatory framework and rules: