Claude Can Now Use Your Passwords Securely

1Password has launched 1Password for Claude, allowing Anthropic’s assistant to complete sign-ins for users without exposing their passwords or one-time codes to the model.

The integration uses what 1Password calls a zero-exposure architecture. Claude requests that 1Password handle authentication, but the credentials are never shown to Claude or loaded into its memory. Before filling in a login, 1Password notifies the user and requests biometric approval.

Once approved, the password manager fills in the credentials and checks whether they were exposed on the page. If the sign-in fails, it clears the entered values and reports the failure back to Claude.

Recommended reading

Florida man charged over Steam game crypto malware

1Password’s Agentic Mode

1Password also introduced Agentic Mode, a feature in its browser extension designed to give users more control when browser-based AI agents take over. When a compatible agent is active, the extension automatically locks and hides its interface. The agent can use only the logins and one-time passwords explicitly approved for the current task.

The company said Agentic Mode works even when the 1Password-for-Claude integration is not configured or 1Password is not needed for the task. It also supports agents other than Claude.

“We need a new security model that is purpose-built for agents, not just humans. The answer isn’t handing agents your secrets. It is to let a user give an agent permission to use a credential without letting the agent see it. Claude knows it used your login; it does not 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.”

The release comes as companies and users debate how much access AI agents should receive. Incidents involving agents deleting email inboxes or disrupting completed work have raised concerns about granting them broad permissions.

×