The government’s Financial Services AI Adoption Plan provides welcome guidance for firms under pressure to adopt AI tools while ensuring consumers remain protected.
Erin Sims, Financial Services Senior Analyst at RSM UK said: “The Financial Services AI Adoption Plan is a welcome step following the Mills Review and reflects the growing recognition that AI is no longer a future issue for financial services, and is already reshaping how firms operate, serve customers and manage risk.
“The UK's existing regulatory framework remains broadly fit for purpose, but the challenge now is how it applies as firms move towards more autonomous AI-enabled decision making. Areas such as Consumer Duty, operational resilience and the regulatory perimeter are likely to come under increasing strain, and firms will be looking for practical guidance on how these requirements should work in practice.
“Consumers are already turning to AI tools for financial guidance, often without understanding what protections exist if the information is wrong or causes harm. As AI takes on a bigger role in financial decision-making, greater regulatory clarity is essential. The proposed review of AI-generated financial guidance is therefore a positive step, and should help address important questions around consumer protection and trust while keeping pace with innovation.
“The opportunities are significant, particularly around productivity, customer experience and fraud detection. However, the same technology is also making fraud more sophisticated, with increasingly convincing scams, deepfakes and synthetic identities creating new risks for firms and consumers alike. If an AI agent is helping customers make decisions, or even acting on their behalf, there must be clarity around responsibility, consent and what happens when things go wrong.
“Boards should already be considering where AI is influencing decisions, who is accountable for its use and whether existing controls remain effective. Consumers may not care whether a decision was made by a person or an algorithm, what matters is whether the outcome is fair, and whether there is clear accountability when things go wrong.”