Major U.S. Banks Block AI Agents from Moving Money Amid Deposit Flight Fears
Major U.S. banks are blocking external AI agents from accessing customer accounts to prevent unauthorized transactions and deposit flight. While Meta's Muse agent gained millions of downloads, institutions like JPMorgan and Citigroup cite security and liability risks. Banks are developing internal tools and restricted read-only access to balance innovation with risk management.
Key points
- JPMorgan, Citigroup, and Bank of America have blocked external AI agents from accessing customer accounts to prevent unauthorized transactions.
- Banks cite security risks, liability concerns, and the potential for AI agents to move idle cash into higher-yielding accounts, threatening deposit stability.
- Charles Schwab is blocking AI agents from moving money but allows read-only access via Model Context Protocol (MCP) servers for research and balance checks.
- JPMorgan is piloting an internal AI tool called Smart Cash to forecast cash flow and maximize yield, aiming to retain deposits without external agents.
- Regulators at the Federal Reserve are concerned about the rise in AI-driven financial transactions and are considering legal clarifications on liability for autonomous actions.
Background
Meta launched its Muse AI agent on September 8, 2026, which quickly gained over 6.6 million downloads. The agent's ability to automate tasks, including canceling subscriptions and managing finances, raised concerns about its impact on bank deposits. Previous coverage highlighted Muse's rapid adoption and the competitive pressure from other AI agents like Instinct and OpenAI's Dots.
How outlets are covering it
Gizmodo emphasizes the low number of attempts to use AI agents for financial transactions, with Charles Schwab CEO Rick Wurster reporting only tens of blocked attempts per week. CNBC highlights JPMorgan's stance on the risks of 'rogue agents' and the potential liability transfer to customers. The Financial Times focuses on the broader impact on bank deposits, noting that JPMorgan, Bank of America, and Wells Fargo hold $1.6 trillion in non-interest-bearing deposits, which could be at risk if AI agents optimize returns. The outlets also differ in their emphasis on the regulatory response, with the Financial Times noting that regulators have privately welcomed the banks' moves but are considering legal clarifications.
Why it matters
The blocking of AI agents by major U.S. banks highlights the tension between technological innovation and financial stability. As AI agents become more autonomous, the potential for unauthorized transactions and deposit flight poses significant risks to the banking system. The moves by banks to restrict access and develop internal tools reflect a broader effort to balance innovation with risk management, which could shape the future of financial services and regulatory frameworks.
What to watch
Banks are expected to continue developing internal AI tools and restricted access protocols to balance innovation with risk management. Regulators may issue legal clarifications on liability for autonomous AI actions, and banks may explore ways to incentivize customers to keep their deposits through loyalty programs and other measures. The ongoing competition between AI companies and banks will likely shape the future of financial services and regulatory frameworks.
- After Muse Launch, U.S. Banks Are Saying Hell No to AI Agents Moving Money Gizmodo
- JPMorgan Chase blocks outside AI agents like Muse, Instinct from directly accessing accounts CNBC
- US Banks Block AI Agents Attempting to Move Money for Customers Bloomberg.com
- Citi becomes latest US bank to restrict AI agents from accessing accounts Financial Times
- ‘You don’t want rogue agents’: Jamie Dimon bars outside AI agents like Meta’s Muse from JPMorgan accounts Fortune
Want the full story? Read the original reporting
Read on Gizmodo