Bank of America warns users against sharing passwords with AI

Bank of America is urging its customer base to refrain from providing AI agents or third-party digital assistants with their personal banking usernames and passwords. In a recent communication titled Protecting your finances in the AI World, the financial institution explicitly warned users against authorizing AI tools to operate on their behalf, a practice that has grown as automated assistants gain mainstream adoption.

Risks of AI account integration

The Scale Report understands that the bank is concerned about the security implications of this trend. If a customer provides credentials to an AI service, they face a heightened risk of unauthorized transactions, account lockouts, or potential data breaches. Security editor Zack Whittaker of TechCrunch notes that these tools could leak private information, effectively handing a golden ticket to cybercriminals.

Accountability in the AI era

Beyond the technical dangers, there is a significant legal component to these interactions. Bank of America confirmed to TechCrunch that the email circulating on social media is authentic. The bank emphasized that per existing account agreements, the account holder maintains full responsibility for any activity initiated through their credentials, regardless of whether a human or an automated bot performed the action.

The challenge for modern banking

This advisory underscores a wider struggle within the financial sector as institutions navigate the integration of agentic payments. While banks and payment providers explore ways to support tools that autonomously execute financial transactions, they are simultaneously grappling with the reality that many unauthorized agents are already attempting to scrape customer data. For now, the safest path remains maintaining strict control over personal login details.