AI-Agent Liability Could Constrain Banking and Commerce Integrations
Reports say major banks place losses from customer-authorized third-party agents, including hacked agents, on customers, while distinguishing their own agent services. A reported Arizona investigation into OpenAI and other firms over possible rogue agents adds a state-level scrutiny signal, but available coverage gives no scope or findings. Google’s call to keep agents within established payment rails and merchant exposure under current chargeback rules underscore unresolved questions around authorization, consumer protection, fraud, and liability.
Sources (4)
Updated Nov 20, 2026