Banks are drawing a line around agentic finance: the UK’s five biggest lenders and major US banks are blocking outside AI agents from customer accounts while the FCA builds rules for agent identity and accountability. Regulators are simultaneously probing how well firms oversee AI after deployment and experimenting with new tools like quantum computing against financial crime.
Top story: HSBC, Barclays, Lloyds, NatWest and Nationwide are reportedly blocking or restricting outside AI agents from customer accounts, with the FCA drafting rules.
UK’s five biggest banks block AI agents from customer accounts
Dealroom News (citing The Telegraph) · Strategy
HSBC, Barclays, Lloyds, NatWest and Nationwide are blocking or restricting AI agents from accessing customer accounts, following similar October moves by US banks. The curbs appear precautionary, with no reported evidence of agents accessing UK accounts, and McKinsey estimates AI-driven switching could shift up to £700 billion of deposits. For practitioners, the dividing line is emerging between AI tools that can view financial data and those allowed to act on it.
FCA builds ‘agentic finance’ framework as banks set own AI-agent rules
Superpower Daily · Regulation
The Financial Conduct Authority is working with the sector and other regulators on rules for AI agents in finance, leaving individual banks to decide access policies for now. Open questions include who is responsible when an agent makes a mistake, and what automated deposit switching does to bank profitability. Compliance teams should expect rules centred on agent identity, delegated authority and accountability.
BoE and FCA survey: GenAI now 56% of UK financial AI systems
Global Regulation Tomorrow (Norton Rose Fulbright) · Risk
The fourth Bank of England and FCA survey finds generative AI is now 56% of reported AI systems, though only 7% are classed as highly material. Most firms expect frontier AI to have a material impact on cyber risk over the next two years. It shows UK finance is using AI heavily for productivity but is still cautious about high-stakes autonomous decisions.
FCA launches world-first regulator-led quantum group to fight financial crime
Crowdfund Insider · Finance
The FCA has set up a Quantum Experimentation Group to test quantum computing for detecting fraud and money laundering, the first regulator-led effort of its kind. Work begins in Q1 2027, with firms invited to register interest by the end of December 2026. Lloyds Banking Group and IBM are among those backing the initiative, which could shape future rules on how such technology is used.
Regulators spotlight gaps in post-deployment AI oversight at financial firms
QA Financial · Legal
Over 90% of firms report pre-deployment review, human approval and vendor due diligence for AI data risks, but only around half report ongoing monitoring, leakage testing or periodic recertification. The piece also notes US supervisors have issued a framework for examining increasingly autonomous systems. The message for risk and compliance leads is that regulators are shifting from approving AI to demanding proof it stays under control.
https://qa-financial.com/november-roundup-of-regulatory-and-compliance-news