More banks say customers using AI agents bear the consequences themselves, even if hacked
More banks say customers using AI agents are on their own — even if they get hacked
Citigroup, the third-largest U.S. commercial bank, revised its terms of service this week to stipulate that customers who use third-party AI tools to manage their personal finances bear all consequences themselves, even if the AI tools leak credentials or are hacked, or hackers use the accounts to commit crimes. A Citigroup spokesperson said the update to the online user agreement is intended to let customers understand the risks of sharing bank credentials or authorizing these tools to access accounts. JPMorgan Chase and Bank of America announced similar policies last month, PNC also revised its terms today, and Citizens emailed customers this week to remind them that they must be responsible for their AI assistants' behavior. By contrast, Anthropic's personal finance tool launched earlier this month is explicitly labeled as read-only and cannot transfer, manage or invest customer funds. The report also noted that banks view AI tools as a competitive threat entering their business, while Citigroup's terms state that using its own AI agent does not carry additional liability.
Several major U.S. banks shift the risk of third-party AI agents operating accounts and hacking losses onto customers, providing industry background for the debate over liability and trust in agent deployment.