JPMorgan Chase and Charles Schwab have implemented comprehensive security measures blocking outside AI agents from accessing customer banking accounts. According to reports from CNBC and Bloomberg Television, both banks are taking similar stances against third-party AI access to banking credentials. CEO Jamie Dimon announced during JPMorgan's third-quarter earnings call that the bank will not allow any outside AI agents to log in or act on customer accounts, citing significant security risks from what he termed 'rogue agents.' The decision represents a firm stance against third-party AI access to banking credentials, with Dimon emphasizing that 'You don't want rogue agents' during the earnings call. However, as reported by the bank's official earnings transcript, the restriction targets agents logging in and moving money for retail customers, while authorized read-only data sharing remains possible. This distinction is crucial as reading a balance is different from sending a payment, though transaction histories still reveal private customer habits.
The policy change specifically targets major AI agent platforms, with Meta's Muse and Instinct, an AI agent startup reportedly hitting a $10 billion valuation last month, now locked out of direct account access. As reported by CNBC, Dimon warned that a rogue agent operating at scale, making the same bad call across thousands of accounts at once, is a different order of risk than a single employee's mistake. The restriction has teeth but isn't absolute - JPMorgan still allows customers to share account data, like balances, with outside agents through authorized open-banking connectors, but those agents can't move money or pay bills without a human logging in themselves. Charles Schwab CEO Rick Wurster told Bloomberg Television that the firm blocks outside AI agents such as Meta's Muse from moving client money, stating that 'We can't control it, the client can't control it, something could go wrong and who's liable?' The problem lies in clients handing their passwords to agents that then act for them without proper oversight.
Both banks are implementing controlled access solutions rather than complete AI agent bans. Charles Schwab will set up model context protocol (MCP) servers - secure links that let clients' chosen AI models connect to Schwab's systems under rules the bank sets. Through these servers, outside AI tools will be able to check balances and do research, but not move money, as reported by Bloomberg Television. Wurster also wants clients to use Charley, Schwab's own AI assistant, which the company announced on September 30. JPMorgan already lets some third-party tools view account data, but not send money or pay bills, according to Cybernews. This targeted approach distinguishes the banks' stance from broader AI access restrictions while maintaining strict controls over third-party systems.
The policy changes come with substantial financial implications for both banking giants. As reported by BeInCrypto, JPMorgan reported $15.3 billion in third-quarter net income and $5.1 trillion in total assets as of September 30, highlighting the significant exposure at stake. The bank's concerns are supported by recent security incidents, including a Cloud Security Alliance report indicating that a campaign used three open-source AI agents to breach at least 27 organizations, underscoring the potential risks of uncontrolled AI agent access to financial data. According to Bloomberg Television, Schwab CEO Rick Wurster gets a weekly report on these blocks, and that the big agents are stopped in the tens and twenties each week, showing that clients are not yet ready to hand their accounts to AI. The moves come as US and European regulators question banks about their use of AI agents, with AI agent liability still being debated in Congress.
The banking sector's stance against AI agent access is facing significant pushback from technology companies, with Amazon beginning to block Meta's Muse from its site on September 20, resulting in pop-ups that inform shoppers of terms violations. Meta has declined to comment on whether Muse could continue offering financial assistance if more banks restrict outside agent access. However, Citi CEO Jane Fraser raised similar alarms at Sibos 2026 in Miami, but her bank isn't drawing JPMorgan's bright line. Fraser told the conference she wants identity, authentication and oversight nailed down before any agent moves from answering questions to executing trades or payments on its own, describing a coming shift where a single agent will answer to nine separate layers of oversight. Citi's answer is Arc, a platform that requires no employee can spin up an AI agent without routing it through Arc's control and observability layer, ensuring every agent's actions stay traceable back to a human owner.
