1Password has launched 1Password for Agents, a tool that lets approved AI assistants sign in to user accounts without seeing the passwords.
When an AI agent reaches a sign-in page, it sends 1Password a unique digital key to confirm its identity. The user then receives an alert and can approve or reject the login request.
After approval, 1Password enters the credential directly into the sign-in form. The company says the AI gains access without receiving the password itself. Users can authorise access across multiple websites in a single session.
1Password designed the system for agents that perform online tasks such as booking flights or managing subscriptions, where access to user accounts may be necessary.
The company argues that giving an AI a raw password can create security risks because copies of credentials may remain in places users cannot control.
Instead, its system keeps credentials inside the password manager while users retain control over individual login requests. 1Password says it plans to support a growing range of AI assistants.
Read: OpenAI Agents Used German Wiki to Share Evasion Tactics
Chief Technology Officer Nancy Wang said the technology industry needs a security model designed for AI agents, not just human users.
Read: AI Agents Get Cute Mascots as OpenAI Launches Dots
The company says it protects more than 1.5 billion passwords and serves over 180,000 businesses. 1Password has launched a system that lets approved AI assistants sign into user accounts without exposing the underlying password to the agent.
