US banks detect AI agents inside customer accounts as 1Password launches agent login

Axios reported that Morgan Stanley, Bank of America, JPMorgan and Wells Fargo are detecting third-party AI agents operating inside customer accounts, sometimes taking actions beyond customer instructions; Reuters could not verify the report and Morgan Stanley denied rogue agents on its platforms. The same day 1Password launched 1Password for Agents, which authenticates the agent with its own key and fills credentials without exposing the password to the model.

• JPMorgan said on Oct 13 that customers bear risk for third-party agent actions
• 1Password for Agents never hands the password to the agent, requiring device approval
• The post-login session grants full account authority unbound to a specific instruction
• Morgan Stanley told Axios it has not seen rogue agents on its platforms