Citibank considers erroneous transactions by external AI granted account access as approved
Original headline씨티은행, 계좌 접근 허용한 외부 AI의 오작동 거래도 승인으로 간주
Not investment advice. Muhabbit scores how much a story may matter, not which way to trade.
Citibank has added a "third-party AI agent" clause to its online terms effective November, treating any access or action by an AI agent as authorized by the customer once login credentials or account access are granted. The bank assumes no duty to investigate whether an agent's trades matched the customer's instructions or whether those instructions were valid. The move shifts liability to customers for errant agent transactions and for account activity stemming from leaked login credentials. It matters because AI agents are beginning to transact directly on customer accounts, and Citibank is among the first major banks to formalize who bears the risk. What to watch next is whether rival banks adopt similar clauses and how customers and regulators respond to the liability shift.
Muhabbit's AI wrote this summary in its own words from TokenPost's report and translated it from Korean. It can get details wrong, so check the source.
Named in this story
Citibank, Citigroup (C)
A linked name opens that asset's sentiment page.