1Password launches 1Password for Agents agent password authorization feature
1Password launches the 1Password for Agents platform, enabling personal AI agents to complete website logins on users' behalf without accessing or storing users' passwords or verification codes. The feature is aimed at individual and family plan users, with initial support for agents including Meta Muse, OpenAI ChatGPT Work and Dots, Anthropic Claude Duet, Grok Bot, and Town. In the flow, the agent first proves its identity to 1Password using a digital key; the user sees on their device the account credentials and task requested for use and approves or rejects; after approval, 1Password directly fills in the username, password, or one-time verification code, and the password always remains in the vault.
Why read this
AI score 58
The original article breaks down the authorization flow in which passwords are not exposed when agents log in, which can be used to assess how to balance AI-run tasks with account security.
Topics
Body · AI translation
Let AI shop, book tickets, and order takeout with peace of mind: 1Password launches agent password authorization feature
English text is being prepared.
English text is being prepared.
English text is being prepared.
The official introduction video is included below:
The feature is available to individual and family plan users, and initially supports mainstream AI agents such as Meta Muse, OpenAI ChatGPT Work and Dots, Anthropic Claude Duet, Grok Bot, and Town.
The company said that as AI agents gradually take on tasks such as ordering food, shopping, booking tickets, and organizing materials, account login has become a key step in their execution of tasks. But if passwords are handed directly to agents, users face the risk of credentials being memorized by the model, leaked, or misused.
1Password's new approach seeks to establish an isolation layer between "authorizing AI to do things" and "keeping passwords secure." The specific process is: after an AI agent encounters a login page, it first proves its identity to 1Password through a unique digital key.
The user then receives a notification on their own device, sees which account credential the agent is requesting to use and for what task, and chooses to approve or deny it themselves. After the user approves, 1Password directly fills the username, password, or one-time verification code into the login form; once login is complete, the agent can continue executing the task, but the password always remains in the 1Password vault.
1Password emphasized that agents do not gain access to the entire password vault, and other passwords not explicitly authorized remain protected by end-to-end encryption. This mechanism is similar to adding a "two-factor verification" layer for AI: the user holds final approval authority, and the agent only receives the temporary login capability needed to complete a specific task.