1Password Rolls Out Tool Letting AI Agents Log In Without Passwords

1Password launched a new product Thursday, called 1Password for Agents, that lets artificial intelligence assistants log into a user's personal accounts without the agent ever seeing the password, according to TechCrunch. The password manager says the tool works with at least five named AI agents at launch and is designed to close a gap that has already drawn scrutiny from the banking industry.

How does 1Password for Agents keep passwords hidden?

The system routes every login through three parties: the agent, the user, and 1Password itself, per TechCrunch's reporting on the launch. When an agent attempts to sign into an account on a person's behalf, it sends a unique digital key to 1Password to verify its identity. 1Password then alerts the user, who must personally confirm the request before anything proceeds. Only after that approval does 1Password fill in the credentials itself, completing the login without ever handing the password to the agent.

Which AI agents support the feature at launch?

1Password says users can choose from Anthropic's Claude Duet, OpenAI's Dots and ChatGPT Work, Grok Bot, Meta's Muse, and agents built by startup Town, TechCrunch reported. The company said it is working with additional partners to expand the list, though it did not name further companies in the announcement.

Why are banks flagging AI agent activity now?

1Password's launch landed the same day Axios reported that U.S. banks have started detecting agentic activity inside customer accounts, according to TechCrunch's account of that reporting. The activity is not necessarily unauthorized; much of it reflects customers deliberately letting automated assistants handle financial tasks. The complication, per bankers Axios interviewed, is that most banks currently cannot distinguish agent actions a customer requested from those they did not, a situation one banker described as "horrifying," TechCrunch reported. A Chase spokesperson told Axios that customers "should not give their usernames and passwords to anyone," a standing policy that predates the rise of AI agents but that the new activity has made more urgent.

What problem does 1Password say it's solving?

In a company blog post cited by TechCrunch, 1Password framed the risk in blunt terms: "Handing over a password, secret, or credential is a common shortcut to an agent, but can leave people with sensitive information in places they can no longer control as a copy gets saved for the next time." The company said the new tool is meant to let people "delegate tasks to personal AI agents while keeping control of the passwords, secrets, and credentials those tasks require."

What does this mean for users weighing AI agent access?

The launch gives consumers a narrower alternative to the current default of typing a password directly into an assistant or copying it into a chat window — a practice security teams have warned leaves a durable, exportable copy of the credential sitting outside the user's control. Whether banks and other institutions adopt compatible verification standards for agent logins remains an open question the TechCrunch report does not resolve.