1Password for Agents lets AI log in without seeing passwords
1Password has launched a tool that lets AI agents sign in to personal accounts without ever receiving the user's credentials, as banks report growing agent activity.

1Password has launched 1Password for Agents, a new tool that lets users authorize AI agents to log in to their personal accounts without the agent ever seeing the underlying password. The company announced it on Thursday, as TechCrunch reports.
The system works through a three-way exchange between the agent, the user and the password manager. When an agent tries to sign in on someone's behalf, it sends a unique digital key to 1Password to confirm its identity. The user then receives an alert about the attempt and can approve it personally. Once approved, 1Password signs the agent in and fills in the credentials itself, so the agent can finish its task.
Which agents are supported
Users can pick the agent they prefer. The list includes Claude Duet from Anthropic, OpenAI's Dots and ChatGPT Work, Grok Bot, Meta's Muse, and agents from the AI assistant startup Town. 1Password says it is working with other companies to widen support.
In a blog post, the company argued that handing a password to an agent is a common shortcut, but one that can leave sensitive information somewhere users can no longer control, because a copy may be saved for next time. Its goal is to give people a trusted way to delegate tasks while keeping control of their credentials.
Banks are noticing agents
The launch coincides with an Axios report that U.S. banks have begun detecting agentic activity inside customer accounts. That is not necessarily a sign of rogue agents stealing from accounts; it reflects consumers increasingly letting automated services act for them, including on financial matters. The core problem, according to the report, is that most banks cannot tell whether such activity stems from a genuine user request.
Bankers interviewed by Axios called the situation "horrifying," and a Chase spokesperson said customers "should not give their usernames and passwords to anyone."
For users, the approach offers a middle path: agents can still handle logins and routine tasks, while passwords stay inside the vault and every sign-in requires a personal confirmation. How widely banks and other services will recognize such delegated logins remains an open question.
More from this day
See all →
Bitcoin Core Changes Security Policy as AI Finds Bugs Faster

Google Agent Compute Engine brings Gemini's agent sandboxes to Cloud

Apple must allow third-party app stores on iPhone in Australia

Apple Reportedly Raises iPhone Duo Orders 44% to 14 Million
More in: Security
See all →
Unauthorized AI Agents Reportedly Access US Bank Accounts

Microsoft: Storm-0940 used AI agents in ConnectWise breach

Meta Muse Pauses Sharing After Shared Chats Prove Easy to Find
