AI in the financial services sector is not a new concept. It has been utilized for various purposes, such as automated stock-trading and credit-scoring algorithms for loan applications, for several years now. However, the rapid growth of AI technology has raised concerns for regulators, particularly those responsible for ensuring transparency and fairness, like the Financial Conduct Authority (FCA). The evolving nature of technology poses a challenge for regulation as it tries to keep pace with technology development. The EU has attempted to address this through the AI Act, a comprehensive set of regulations, while the UK takes a more principles-based approach, allowing individual regulators like the FCA to respond to AI on a sector-by-sector basis. Nevertheless, formal regulatory responses are still in progress. The FCA is cautious about setting fixed rules that could hinder innovation but recognizes that AI can also amplify existing inequalities, posing a dilemma for regulators. FCA's CEO, Nikhil Rathi, has pledged a pro-business approach and plans to open up the AI "sandbox" for testing non-compliant products. The FCA's concern lies in consumer-facing financial services that use non-traditional data, like social media behavior and shopping habits, to assess loan applications. They worry about potential consumer harm resulting from biased or inadequately processed datasets.

Explainability is also a persistent issue, as businesses need to be able to justify their AI models' decision-making process. The balancing act for regulators lies in promoting competition and seizing opportunities while protecting consumers and addressing public concerns. The FCA is careful in proposing formal regulations due to the dynamic nature of the industry. Instead, they favor principles-based guidance, aligning with the UK's broader approach to AI regulation. Both regulators and businesses face challenges in keeping up with the fast-moving landscape of AI. Start-ups, with their "move fast and break things" mentality, need to navigate compliance requirements that may not align with their practices. There is a possibility that regulators may require designated individuals within organizations to take responsibility for AI, expanding the existing UK Senior Managers and Certification Regime. However, this proposition would require substantial government support. The FCA will rely on existing principles for businesses to guide AI and machine learning activities, without requiring a fundamental overhaul of financial services regulation.