UK Financial Regulator Warns of AI Risks and Benefits in Financial Services

AI risks making some people 'uninsurable', warns UK financial watchdog

Consumers risk becoming "uninsurable" due to artificial intelligence systems in the sector, the head of Britain's financial watchdog has warned, despite calling for the insurance industry to be more "willing to experiment" with new technology. "We want safe and responsible use of AI to drive beneficial innovation," said Nikhil Rathi, chief executive of the Financial Conduct Authority, in a speech on Thursday. "But also an open conversation about the risks and trade-offs." Giving the example of "AI-enabled hyper-personalisation of insurance", Rathi said this could benefit many consumers but also warned that it "runs the risk of rendering some customers 'uninsurable', or even potential discrimination". Some experts have raised concerns about AI use in areas such as health insurance, where live data could increase personalisation and lower costs for some consumers but also risks making it harder for some unhealthier people or those without access to technology to get affordable cover. Eiopa, the EU insurance regulator, said a few years ago that companies should "make reasonable efforts to monitor and mitigate biases from data and AI systems", given the risk that algorithmic pricing models could end up discriminating against certain people. The FCA launched a discussion paper on AI two years ago. While it has introduced few specific rules governing its use by financial services providers, it said in April that it would "continue to closely monitor" adoption of the technology. Rathi equated the risks of AI usage with the recent controversy over the dynamic pricing model that caused a surge in the cost of highly sought-after tickets for concerts by music group Oasis. "Just because something can be done, doesn't necessarily mean the public will accept it," he said, adding: "Not everyone will 'Roll With It'." The FCA chief, who has focused on protecting consumers since joining the regulator in 2020, offered encouragement to those who think its vast rule book hampers innovation by saying it should be "ready to rethink some of our rules and regulatory approaches" to boost financial inclusion. He acknowledged there were "tensions between the prescription previously necessary, and what a fast-digitising financial services market now requires". Emphasising the link between greater economic growth and increased financial inclusion, he said Singapore's position at the top of financial inclusion rankings showed how "a successful global financial centre and financially inclusive economy can go hand-in-hand". He added that the UK came seventh in the latest ranking and there were still 1.1mn people in the country without a bank account. He cited several examples of how technology had been used successfully to boost financial inclusion in other countries, including Brazil's instant payment service Pix and India's biometric ID system Aadhaar. But he also gave some UK examples of innovation, including Finexos, which is working on availability of affordable credit, and Noggin, which produces an alternative credit score for consumers with limited credit history. The previous Conservative government gave the FCA a new secondary mandate last year that required it to take into account the impact of regulation on growth and competitiveness. On Thursday, the FCA bowed to pressure from the UK's £267bn investment trust industry by announcing its members would be exempt from disclosure rules that stem from EU legislation and are due to be replaced next year. He said the regulator "accepts that the risk of a few experiments failing or some people not benefiting from innovation, is outweighed by the potential benefit to the majority of consumers, and long-term growth and productivity improvements. "Let's open up this debate and be willing to experiment and learn," he added. "And that includes experimenting with our processes and rules."

FCA Head Nikhil Rathi Says AI Has Benefits and Risks for Financial Inclusion | PYMNTS.com

The hyper-personalization of insurance enabled by artificial intelligence could benefit some customers but render others "uninsurable," Nikhil Rathi, CEO of the United Kingdom's Financial Conduct Authority (FCA), said Thursday (Sept. 19). In a speech drafted for delivery at StepChange Connected 2024 in Leeds, Rathi cited this as an example of how regulators must consider both the risks and the benefits of AI and other technology when promoting financial inclusion. "We want safe and responsible use of AI to drive beneficial innovation," Rathi said. "But also an open conversation about the risks and trade-offs." Other examples of technology risks include the privacy concerns associated with better data sharing and the recent controversy around the dynamic pricing used in a sale of Oasis concert tickets, Rathi said. Digital inclusion can also promote financial inclusion, Rathi said. He cited as examples the efforts of Finexos to increase the availability of affordable credit, the work of Noggin to produce an alternative credit score for consumers with limited credit history, and a report from an unnamed FinTech that anonymous chatbots providing debt advice benefit consumers by reducing the stigma associated with debt. "Do we accept that the risk of a few experiments failing or some people not benefiting from innovation is outweighed by the potential benefit to the majority of consumers and long-term growth and productivity improvements?" Rathi said. "I freely admit that we don't yet have full estimates of the benefits of inclusion to growth and productivity. But experience elsewhere suggests that resolving foundational issues could have big impacts." Rathi suggested that financial inclusion can be promoted by improving financial literacy, enhancing digital literacy and inclusion, embracing digitalization and technological innovation, and "being more ready to experiment and take some measured risk to deliver better long-term economic outcomes." It was reported in June that financial services companies have been slow to embrace AI products because of regulatory concerns and worries about job losses.