America's biggest banks are facing an unusual security threat where artificial intelligence (AI) agents can enter customer accounts using credentials and, in some cases, appear to take actions without clear instructions from the account holder, according to a report by Axios.

The growing use of AI agents, which can carry out tasks on behalf of users, is creating a security challenge for financial institutions. Banks are struggling to distinguish between human customers and automated agents, raising concerns about potential unauthorised transactions, fund transfers and other actions that could affect customers' finances.

The problem has prompted several major banks to restrict third-party AI agents from accessing their platforms, while financial institutions work to strengthen their security measures and revise their terms of service to address the emerging risks.

MAJOR BANKS MOVE TO RESTRICT AI AGENTS Banks including Morgan Stanley, Bank of America, JPMorgan Chase, Wells Fargo and Citi are working to contain the risks associated with third-party AI agents, people familiar with the issue told Axios.

In many cases, the agents access customers' accounts using credentials provided by the account holders themselves. Although customers may authorise such access, the practice violates the guidelines of most banks, which require people to to their own accounts.

The concern extends beyond access to financial information. Banking executives fear that AI agents with permission to make purchases or move money could take actions with direct financial consequences for customers.

Banks have also tracked instances in which AI agents, programmed to fulfil requests or complete tasks on behalf of users, contact customer service call centres or appear to act independently of user instructions, the report said.

MORGAN STANLEY, JPMORGAN AMONG BANKS TIGHTENING ACCESS Several financial institutions have begun restricting third-party AI agents as they assess the potential threats.

Morgan Stanley has started blocking third-party agents from accessing its platforms, according to sources familiar with the decision. JPMorgan Chase, Bank of America and Citi are also restricting such agents.

A Chase spokesperson advised customers against sharing their usernames and passwords with anyone.

A Morgan Stanley spokesperson, however, said the bank had not observed instances of "rogue AI agents accessing our platforms".

Bank executives have described the situation as deeply concerning, particularly given the possibility that agents with access to customer accounts could take actions affecting their money.

WHY AI AGENTS POSE A SECURITY CHALLENGE One of the biggest difficulties for banks is determining whether activity on a customer account is being directed by the account holder or carried out independently by an AI agent.

According to Axios, some banks are struggling to distinguish between different agents and establish whether an agent has been authorised by a customer or has gained access without permission and gone rogue.

This uncertainty complicates efforts to prevent potentially harmful activity before it affects customers. Even when an account holder initially authorises an agent, banks may find it difficult to determine whether subsequent actions remain within the scope of that permission.

The rapid development of AI technology is adding to the challenge, exposing new weaknesses in real time and making it difficult for financial institutions to respond as quickly as the risks emerge.

BANKS EXPLORE READ-ONLY ACCESS AS A MIDDLE GROUND Despite their concerns, banks recognise that customers will increasingly want to use AI agents to manage financial tasks. Several institutions have therefore authorised, or are moving towards allowing, read-only access.

This arrangement would enable AI agents to view customers' financial information without being able to initiate transactions or otherwise take action on their accounts.

Banks already permit similar access for third-party applications such as Quicken, which customers use for budgeting and other financial activities.

Industry and banking executives told Axios that financial institutions understand the need to find a middle ground that allows customers to benefit from AI agents without exposing their accounts to unacceptable risks. For now, however, executives believe the dangers of allowing unrestricted access outweigh the potential benefits.

BANKS PREPARE FOR WORST-CASE SCENARIOS Financial institutions are investing in measures to prevent AI-related security incidents, but they are also preparing for situations in which their safeguards fail.

According to sources cited by Axios, some banks are planning for the immediate aftermath of a potential incident, including ways to quickly secure accounts or restore customer access if an AI agent causes a disruption.

Banks have significant financial resources and experience managing critical infrastructure, which puts them in a strong position to respond to emerging AI threats. However, one source told Axios that money alone might not be enough to resolve the problem, given the speed at which the technology is evolving.

As AI agents become more capable of carrying out tasks on behalf of users, banks face the challenge of allowing legitimate uses of the technology while preventing automated systems from compromising customer accounts or taking actions that users never intended.