UK Finance Leaders Warn of Rising Risks as AI Adoption Outpaces Governance

Senior figures across the UK financial services sector are warning that the country is falling behind global peers in establishing a shared, practical standard for governing AI — a gap they say is already creating systemic risk as adoption accelerates.

A new report from compliance technology firm Zango AI, based on interviews with leaders from Lloyds, Monzo, Stripe and others, highlights growing concern that firms are being forced to design their own governance controls in isolation. The findings land as the Bank of England prepares to convene the Treasury, FCA and National Cyber Security Centre to assess risks posed by Anthropic’s Mythos model.

While the US and Singapore have both issued detailed, sector‑specific AI governance frameworks for financial services this year, no equivalent exists in the UK or EU. The US Treasury’s February framework involved 108 financial institutions and input from agencies including NIST; Singapore’s Monetary Authority followed with its own guidance in March. Zango argues that the absence of a UK standard is leading to inconsistent controls, duplicated effort and oversight gaps that could be exploited at scale — precisely the risks regulators are now scrutinising.

The report calls for a practitioner‑built, regulator‑supported implementation standard modelled on the Joint Money Laundering Steering Group, which provides industry‑developed guidance for financial crime compliance. In a foreword, Lord Clement‑Jones warns that the UK must not wait for “the first major AI‑fuelled financial scandal” before translating high‑level principles into operational practice.

Leaders interviewed for the research say the nature of AI systems now being deployed — generative and agentic models producing context‑dependent outputs that cannot be fully validated in advance — is outpacing existing governance frameworks. Risk and compliance teams are struggling to keep up, with some institutions unable to identify all AI tools in use across their own organisations.

Criminal groups are already exploiting the gap: global fraud losses reached $579 billion in 2025, with 90% of financial professionals reporting an increase in AI‑enabled attacks. Dean Nash, adviser to Zango and global COO (legal) at Santander, says current governance models were built for predictable systems, not tools that “make judgements” and behave differently across contexts. “Most firms are trying to solve it alone, without a shared standard to work from,” he notes.