More banks says customers using AI agents are on their own — even if they get hacked
First reported by TechCrunch ·
If you use an AI agent to manage your bank account, you are now personally liable for all resulting losses, even from hacks.
Citibank has updated its terms of service to state that customers are solely responsible for any financial losses incurred when using third-party AI agents to manage their bank accounts. This policy extends to losses resulting from hacks or breaches of these AI tools, even if customer credentials are leaked. This stance mirrors similar policies recently announced by JPMorgan Chase and Bank of America, the two largest U.S. commercial banks, and also by PNC Bank. Financial institutions are increasingly wary of integrating unreliable AI actors into their systems and are pushing back against assuming liability for automated transactions or security failures initiated through these external services. Several banks view AI tools that facilitate easier consumer access to financial services and potentially bypass traditional banking channels as a competitive threat.
Major U.S. banks are pushing back against the integration of third-party AI agents into personal finance management, establishing policies that absolve them of responsibility for customer losses. This move signals a growing tension between financial institutions' desire to maintain control over customer relationships and the burgeoning market for AI-powered financial tools. Banks appear to be prioritizing the protection of their existing revenue streams and customer relationships over embracing potentially disruptive AI technologies that could facilitate easier consumer access to competitive financial products. The stance taken by major players like Citi, JPMorgan Chase, and Bank of America suggests a unified front against delegating account access and transaction authority to external AI services.
This collective policy shift by banks indicates a significant hurdle for AI startups aiming to enter the personal finance sector, especially those seeking to offer transaction capabilities. Companies like Anthropic are opting for read-only AI tools to circumvent these liability concerns, highlighting the current limitations for more integrated AI financial assistants. As banks fight to control customer pathways and revenue, the future of AI in personal finance hinges on whether these institutions will develop their own secure, integrated solutions or if external providers can navigate the escalating trust and liability challenges to offer robust, accountable services.
AI-written summary. May contain errors.