1Password lets AI agents sign in without seeing passwords
1Password for Agents requires user approval for each sign-in and fills credentials without handing them to the agent.

AT A GLANCE
Key points
- The agent never sees or retains the password.
- Each sign-in requires approval on the user's own device.
- The system is designed for multiple personal-agent platforms.
What is changing?
1Password has introduced 1Password for Agents, a system that allows AI agents to work inside user accounts while keeping passwords hidden from the agents. When an agent reaches a sign-in form, it identifies itself with a unique digital key and 1Password sends the user a request explaining which account is needed and why.
If the user approves the request on their own device, 1Password fills the credentials directly into the form. The agent can then continue the task without seeing or retaining the password. Requiring a separate approval for each sign-in is intended to keep access scoped to a specific task.
The company says the system is designed to work with personal agents including Meta Muse, ChatGPT Work and Dots, Claude Duet, Grok Bot, and Town. Support is also being developed with Browserbase, Browser Use, Dana, Instinct, Kernel, Perplexity, Ramp, Tomo, and Wajo.
The announcement arrives as agents begin handling shopping, bookings, and account management. Instead of placing credentials inside the agent's context, the design keeps human approval and a separate trust layer at the center of the access flow.
Explore the original sources
Source date: 22 October 2026.
Published by AI Radar: 22 Oct 2026, 17:11 (Istanbul, UTC+3).
FOLLOW THE TOPIC
Topics and tags
Share your discovery.
Send this story to a friend.
Reader comments
What do you think about this development? Share your experience or question. Comments are published after editorial approval.
Your name/nickname and comment become public if approved. Read: Privacy notice.






Loading comments…