1Password Lets AI Agents Log In Without Seeing Passwords

The company said the service is designed as a security layer between autonomous AI agents and user accounts. It arrives as consumers increasingly use agentic AI for tasks such as online shopping and booking travel. Until now, users had to sign in manually when an agent reached a login page, interrupting the handoff of tasks to AI agents.

Under the system, an agent identifies itself to 1Password with a unique digital key. The user must still approve the login request on a trusted device, while password sharing is handled securely in the background. 1Password said plaintext credentials and one-time codes are not shared with the AI agent, reducing concerns about rogue misuse.

“An agent should be able to do the job it’s been asked to do without getting permanent access to everything behind the account,” Chief Product Officer Aashish Tripathi said.

1Password said it is working with AI companies including OpenAI, Anthropic and Meta on tighter integrations. Grok Bot and Meta Muse are among the first agents to receive the rollout over the coming days, while ChatGPT Work, Dots and Claude Duet are listed as coming soon.

Individual and Family customers using 1Password.com hosting will receive the upgrade for free, the company said. Customers on 1Password.eu and 1Password.ca hosted plans will not.

TechRadar reported that banks have been concerned about unauthorized AI agent access to customer accounts.