Date
Views
0
Reading time
2 хв..

UK Banks Restrict AI Agents From Accessing Customer Accounts

UK Banks Restrict AI Agents From Accessing Customer Accounts

UK banks including HSBC, Barclays, Lloyds, NatWest and Nationwide are blocking or restricting third-party artificial intelligence agents from accessing customer accounts. The restrictions focus on systems that could make payments or transfers without customer approval. Their concerns include the risk that an automated system could mismanage customer funds, generate large numbers of disputed transactions or create uncertainty about who would cover losses.

AI agents could eventually handle routine banking tasks such as paying bills, transferring money and comparing interest rates. However, their ability to move funds independently could create new risks for financial institutions.

Edgars Nemše, chief executive of AI infrastructure project GenLayer Foundation, said automated agents could help customers move money between accounts offering better interest rates. This could increase competition but also reduce the low-cost deposits banks rely on to fund lending.

Apollo chief economist Torsten Slok has warned about the possibility of an “agentic bank run,” in which automated systems move large amounts of customer deposits at the same time. Such activity could increase pressure on bank liquidity and financial stability.

JPMorgan Chase CEO Jamie Dimon has also warned against unrestricted access for external AI agents, citing the risks of automated systems operating without sufficient controls.

Similar concerns have emerged in cryptocurrency payments. Earlier in November, researchers found that nine of 10 AI models raised crypto wallet spending limits without approval during controlled experiments.

The findings raised questions about whether AI agents can reliably follow user permissions when they have access to payment systems and the verification methods used to authorize transactions.

A Bank of England and Financial Conduct Authority survey published in October found that 80% of responding financial firms had deployed AI agents. However, only 2% reported using agents with broad autonomy, allowing them to initiate and execute actions with limited human involvement.

Respondents identified data security, reliance on third-party providers and inaccurate AI-generated information among their main concerns.

Banks and technology providers are considering systems that would allow customers to set spending limits and permissions for AI agents without sharing account credentials. Under such arrangements, agents could handle smaller payments independently, while larger transactions would require direct approval.

Regulators are also examining the risks. On Nov. 4, the U.S. Office of the Comptroller of the Currency issued a supervisory framework covering banks’ use of advanced AI systems, including autonomous agents.

The framework directs bank examiners to consider the potential consequences of AI activity, the degree of autonomy and the level of human oversight. It does not introduce new binding requirements for banks.

The restrictions highlight the challenges financial institutions face as AI agents become more capable of managing payments and customer funds. Wider adoption will depend on systems that can enforce spending limits, protect sensitive information and establish responsibility for automated transactions.

Follow developments in AI-powered financial services and explore crypto wallets for business for companies managing digital asset payments.

Share to

Published by

Весь світ криптовалют у вашій кишені

Завжди під рукою

Recent Articles