Stay informed with free updates
Simply sign up to the UK financial regulation myFT Digest — delivered directly to your inbox.
The UK’s top financial watchdog has taken a completely different approach to overseeing companies’ use of technology such as artificial intelligence, committing not to “come after you every time something goes wrong”.
Nikhil Rathi, chief executive of the Financial Conduct Authority, said: “We have to think totally differently, not just about risk in the system but the way we regulate and how our organisations interact together.”
Speaking at the FT Global Banking Summit on Tuesday, Rathi said the FCA had decided against making new rules for AI in financial services because “the frontier of that technology is moving every three to six months”, meaning its usual approaches “just don’t work in that kind of environment”.
“There needs to be a different relationship between regulator and regulated,” Rathi said. “We are not going to come after you for everything that goes wrong — what we will be concerned about is egregious failures that are not dealt with.”
The FCA boss pointed to the EU’s recent announcement that it was pausing parts of its landmark AI legislation as a warning sign for other jurisdictions.
“I’m not sure what detailed rules around AI would look like,” he said. “I think we have to be quite careful. If you look at where that has been done elsewhere, say, for example, in the EU they put the AI Act in place but then a couple of years later, just as it was becoming effective, they had to pause elements of it because the world had just moved on.”
The FCA and other UK financial regulators are under heavy pressure from ministers and City of London executives to ease the burden of rules to improve the country’s competitiveness and revive its flagging economic growth.
Rathi said weak UK productivity growth, rapid technological advances, rising defence and security threats and a global shift by regulators to modernise their rules had prompted the FCA to take a fundamentally different approach.
“The environment has changed; therefore, we need to change,” he said. “If we do not become rapid adopters of new technology, how are we going to go after that productivity challenge that we have? Go and innovate — we want you to use that technology.”
Financial services companies were using AI to transform many areas of their operations, including fraud detection, customer service, algorithmic trading and corporate research, Rathi said.
Speaking at the banking summit on Wednesday, Matt Weaver, head of solutions engineering, Emea at OpenAI, unveiled a series of partnerships with UK financial companies, including neobank Revolut and private equity firms Hg and Permira.
Allica Bank, one of OpenAI’s newly announced partners, said the start-up’s AI models had allowed the business banking fintech to save an average of 20 minutes when processing lending applications. Revolut said it was using OpenAI to strengthen its risk and compliance controls.
“All of that is good stuff, which we want to encourage and don’t really want to get in the way of by putting new rules in place,” the FCA boss said, adding that the regulator accepted “there will be bumps in the road” with innovation.
Rathi said the risks of new technology were “quite sobering”, particularly given the “doomsday scenarios” for the financial system he had been told about that stem from the potential consequences of advances in quantum computing.
However, he said the consumer duty rules introduced by the FCA a couple of years ago — requiring companies to ensure their customers have a good outcome without specifying exactly what that was — meant it did not have to write new rules governing how companies use AI.
Separately, the FCA said on Wednesday that the first group of companies — including NatWest, Monzo, Santander and Scottish Widows — had joined its AI live testing initiative, which provided them with “tailored support” on how to use the technology “safely and responsibly”.
Preetham Peddanagari, UK financial services technology consulting leader at EY, said the FCA initiative was “very welcome”, adding that a recent EY survey had found a quarter of companies using AI “still lack basic controls to protect customers and ensure compliance”.