A lack of clarity around AI regulation is having a “chilling effect on high-end AI adoption” in financial services, MPs have been told by industry.
The UK’s market watchdogs are risking serious harm to the sector through their hands-off approach, the Treasury Committee warned as a result.
The Committee said the Financial Conduct Authority (FCA), Bank of England (BoE) and HM Treasury have taken a “wait-and-see” approach to AI and failed to exercise new powers to scrutinise AI and cloud vendors.
"On the hook"
In a January 20 report, MPs noted that David Geale, the FCA’s Executive Director for Payments and Digital Finance, said existing rules meant individuals were “on the hook” for harm caused to consumers through AI.
Innovate Finance, a trade association, argued that senior managers struggled to assess this risk, arguing that the “lack of explainability” of AI models directly conflicted with the regime’s requirement for senior managers to demonstrate they understood and controlled risks.
(Geale hit back, saying “‘I did not understand it’ is not a defence, because you should understand what you are deploying, or you should understand what you are seeking to achieve” and denied new rules are needed.)
"Little practical clarity"
In a report published January 20, the committee said: “We recognise that it is difficult to provide prescriptive regulation in the context of fast-moving technological change. However, the current approach gives firms little practical clarity as to how existing rules apply to the use of AI.”
Many organisations submitting evidence to the committee agreed.
Law firm Burges Salmon LLP told MPs: “The regulatory environment may limit braver adoption until overlaps and inconsistencies in laws and/or regulation can be resolved…”
See also: UK dithers on tech’s “critical third parties”
The FCA and BoE both say current regulations are adequate for AI.
Jonathan Hall, a member of the Bank's Financial Policy Committee, told MPs a "technology agnostic" approach helped regulators remain flexible amid fast-moving development.
"Enough regulatory bite"
Jessica Rusu, the FCA’s CDIO, agreed. The existing Senior Managers and Certification Regime gave the FCA “enough regulatory bite that we don’t need to write new rules for AI” she said.
But during a hearing with regulators, committee Chair Meg Hillier questioned the institutions' ability to demonstrate a clear understanding of the specific risks of certain technologies or between AI models.
AI champions?
The Treasury says it is encouraging the adoption of AI tools in the financial sector to support wider government growth plans.
On Tuesday, the Treasury added two banking tech leads, Starling CIO Harriet Rees and Lloyds Head of AI Rohit Dhawan, to its AI Champions scheme.
They report to Economic Secretary Lucy Rigby and support safe and swift AI adoption in the financial sector.
Appetite for AI regulation
During inquiries, the committee heard demand has been high for the regulator’s AI-specific tools. Jessica Rusu, the FCA's Chief Data, Information and Intelligence Officer, said 132 firms had applied to use the FCA’s new NVIDIA-powered supercharged sandbox in 2025, and around 50 had applied for its AI live testing sandbox.
Rusu said she had “doubled the size” of her team in four years to deal with the demand.
But a lack of specific AI regulations had left banks with "little practical clarity" on where current rules applied, the committee contended.
It urged the FCA to publish "comprehensive, practical guidance" on the use of AI by the end of the year, introduce AI-specific scenarios into their stress-testing programme, and launch a voluntary test for financial services to model the impact of specific situations, such as a major cyber attack.
Third-party reliance
The committee also said new “critical third-parties regime” powers, allowing the government to bring tech suppliers under the remit of financial regulators, should be flexed by the end of 2026.
A letter from Economic Secretary Lucy Rigby in November said the department expected to make the first designations within 12 months. AWS and Google Cloud are among those expected to be subject to the rules.
At a meeting of the BoE and FCA’s AI Consortium in December, members acknowledged an “increasing reliance on a small number of third-party AI providers” but did not identify ways to overcome the issue.
