Citi holds customers who use third-party AI agents responsible for the agent’s actions and any attacker’s
Citibank updated its Online User Agreement in November 2026 with a new section on third-party AI agents.
Sources: Citi, TechCrunch
In this story
Citibank updated its Online User Agreement in November 2026 with a new section on third-party AI agents. Under it, if a customer gives login credentials to an agent, Citi treats all of the agent’s access and actions as authorized by the customer. If those credentials are leaked or passed on to others, actions taken with them are also treated as authorized by the customer. The customer is likewise responsible for an agent misreading instructions or exceeding its authority, except as applicable law permits. The agreement excludes AI tools Citi itself provides. According to TechCrunch, Citizens sent customers a similar email and PNC changed its terms on Nov. 18; JPMorgan and Bank of America announced similar policies last month.
Sources
4 sources
- “Citi® Online User Agreement”Citi’s Online User Agreement, updated in November 2026; the new Section H covers third-party AI agentsciti.com
- “Citibank Online - Agreement for Online Funds Transfers”Section 4 of the funds transfer agreement: instructions from a third-party AI agent are deemed authorized by the customeronline.citi.com
- “More banks say customers using AI agents are on their own — even if they get hacked”Citi’s statement, PNC and Citizens notices, and JPMorgan’s and Bank of America’s policies from last month · techcrunch.com
- “Bank of America kindly asks that you don’t share your banking password with your AI assistant”Bank of America’s customer email and warning · techcrunch.com
About this story
- Written by
- Yiğit Kasap · JARRUS
- Story date
- On the site
- since Nov. 18, 2026
- @jarrus.tech posted this story on Nov. 18, 2026.
- Short link
thejarrus.com/en/citi-ai-agent-liability- Report an error
- Spotted an error in this story? [email protected] · Instagram