UK financial regulator weighs oversight for general-purpose AI

Sheldon Mills, Executive Director at the FCA, warned that the current rulebook is insufficient as financial firms increasingly integrate these technologies into customer-facing operations. While many institutions initially restricted AI to back-office tasks, the shift toward using these models for investment guidance and complaint handling has exposed a critical gap in consumer protection. Users often mistakenly assume that AI-driven financial tools carry the same safeguards as traditional regulated services, a misunderstanding that could lead to significant financial harm.

Beyond consumer protection, the regulator is concerned about systemic concentration risk. With 81% of financial firms globally adopting AI, the industry is becoming dangerously dependent on a narrow group of technology providers. This shared reliance on a few cloud infrastructures and large language models creates potential single points of failure, raising the specter of correlated market behavior and widespread instability. The FCA intends to spend the next three to six months evaluating whether to expand its regulatory perimeter to bring these general-purpose tools within its scope, ensuring that the financial sector’s digital infrastructure remains resilient as it evolves.