Claude can now enter all your passwords for you – if you give it permission

1Password Unveils Claude Partnership That Lets Anthropics AI Authenticate On Users’ Behalf Via Zero-Exposure Architecture
Users authenticate each login with biometrics
New agent mode in the browser extension locks the interface when AI agents take over

Top password manager company 1Password has launched a new tool that allows artificial intelligence assistant Claude to authenticate on behalf of their user, thus completing tasks that were previously impossible without major security trade-offs.

1Password for Claude is built on “zero-exposure architecture” – so in practice this means that Claude can essentially ask 1Password to complete the login process, but it will never see the credentials and they will never be loaded into its memory.

In return, 1Password will notify the user and request biometric authentication before proceeding. Once given, it will autofill the credentials and check if they were exposed on the page or not. If the submission fails, it will clear the populated values and report back.

Agent mode

“We need a new security model that’s purpose-built for agents, not just people,” said Nancy Wang, CTO of 1Password. “The answer is not giving agents your secrets. It’s letting a user give an agent permission to use a credential without letting the agent see it. Claude knows it used your login; it doesn’t need the password or one-time code in its context. That distinction is where trust in agents starts and the foundation we’re building with Anthropic.”

To further strengthen its security position, 1Password also announced Agentic Mode, a new feature in the browser extension that gives users visibility and control over browser-based AI agents. When a compatible AI agent takes over, the 1Password extension automatically locks and hides the interface. The agent can only use the logins and OTPs that are expressly approved for the current task.

Even if the integration isn’t set up, and even if 1Password isn’t required for the current agent task, Agent Mode works, the company emphasized. Other agents, besides Claude, are also supported.

Currently, a larger debate is underway about how many permissions AI agents should receive and under what rules. We’ve already seen horror stories about AI agents deleting people’s entire email inboxes or otherwise ruining days of hard work. Whether or not this increases, or whether most people remain skeptical of allowing AI access to certain services, remains to be seen.