Banks in the U.K. are bearish on new forms of artificial intelligence, leading the country's regulator to look for ways to make the technology's usage more amenable.
The Financial Conduct Authority will host a discussion with banking industry executives in May ahead of changing AI compliance requirements. The move follows a
That's created competitive pressure as
"These survey results appear to demonstrate a lack of confidence amongst some firms to develop and adopt AI technology, as well as potential uncertainty around the interactions between our regulatory regimes," the FCA said in a statement on the Bank of England survey.
The challenge for regulators is balancing the speed of new technology with the need to protect consumers from fraud, data privacy and other
While all C-suite bank executives in the U.K. said they are "AI-ready" strategically, 43% have a "well-developed AI technology plan," according to a survey by
"At
In the U.S., banks are subject to existing Federal Reserve rules governing general data algorithms that are not necessarily specific to AI. In the EU, the
Bank adoption in the U.S. has been constrained less by consumer rules and more by concerns about fairness because AI models may have embedded biases, and so usage is focused more on customer support and employee efficiency, McPherson said.
The concerns about AI regulation in Europe extend beyond the U.K., he said.
"At the recent AI Action Summit in Paris, there was a definite shift in priorities from safety to speed," McPherson said. "Europeans see what is happening in the U.S. and China, and are worried about falling behind."
The May U.K. summit will address rules such as the General Data Protection Regulation, or GDPR, which places limits on data collection, storage and processing. And the Consumer Duty requires companies to demonstrate that AI and other technology provides "good" outcomes for consumers.
While these rules are not likely to be revoked or substantially weakened, the FCA — with input from stakeholders — will likely publish new guidance on how to develop AI-powered bank products in line with existing rules.
"AI moves faster than the regulators," Elizabeth Gujral, a director at Cornerstone Advisors, told American Banker. "If you don't have clear regulations, you could have flawed or biased AI models. Or, from the bank side, the lack of clear rules could cause banks to fall behind."
In a Wednesday email on the May summit, the FCA told American Banker, "We want to enable a safe and responsible environment for the use of AI in U.K. financial markets, in a way that drives innovation that benefits consumers and markets and supports growth and competitiveness of financial services.
"We will continue to rely on existing frameworks and avoid introducing additional regulations for AI. But we continue to engage closely with industry to ensure they have the clarity they need and that we adapt to the speed, scale and complexity of AI."
Another topic the FCA will address is how to govern individual usage at the corporate level. Banks are also challenged to manage compliance while staff use AI programs on their own devices.
"One bank told me you'd be very naive to think that someone isn't using AI in pretty much every area of your bank today," said Gareth Lodge, a senior analyst at Celent, who hosted a recent bank panel on AI. "Banks both want and need to manage and control the use of AI, for many reasons, but not least the regulatory consequences if they get it wrong."
