Citi makes customers responsible for losses caused by third-party AI agents, even after hacks

Citi's updated online user agreement follows similar policies announced last month by JPMorgan and Bank of America.
Why it matters
Personal-finance AI agents can be given bank credentials to analyse spending, offer advice and make payments. Citi, the third-largest US commercial bank, now says customers who grant such agents account access are responsible for activity conducted through them, including when they are hacked. A Citi spokesperson told TechCrunch the change aims to clarify the risks of sharing banking credentials with third-party AI tools. Banks have said they will not be responsible for reversing transactions made by automated assistants.
Key facts
- Citi updated its Online User Agreement this week
- Customers responsible for activity by third-party AI agents
- Applies even if the agent is hacked
- JPMorgan and Bank of America announced similar policies last month
- Citi is the third-largest US commercial bank