1Password for Agents lets AI agents log in without ever seeing your password

1Password has launched a platform that lets AI agents log into personal accounts on a user's behalf while keeping the actual credentials out of the agent's hands. According to TechCrunch, the product, called 1Password for Agents, arrived on Thursday as security vendors begin building answers to the access consumers are handing to automated assistants.

How it works

The system sets up a three-way exchange between the AI agent, the user, and the password manager, TechCrunch reports. When an agent tries to sign in, a unique digital key is sent to 1Password to verify the agent's digital identity. The user then receives an alert about the login attempt and confirms it personally. Only then does 1Password complete the sign-in, entering the stored credentials on the agent's behalf so it can carry out the task at hand.

The core idea is that the agent gets authenticated and authorised without ever receiving the secret itself. 1Password argued in a blog post that handing an agent a password, secret, or credential is a common shortcut that can leave sensitive information in places people "can no longer control," since a copy tends to be saved for reuse the next time.

Which agents are supported

At launch, users can choose their assistant: Claude Duet from Anthropic, OpenAI's Dots and ChatGPT Work, Grok Bot, Meta's Muse, or agents from Town, an AI assistant startup. 1Password says it is working with other companies to expand the service.

Banks are already seeing agent traffic

The launch lands amid mounting unease about agents acting inside consumer accounts. TechCrunch notes that the product shipped on the same day Axios reported that U.S. banks have started detecting agentic activity within user accounts.

This is not necessarily a wave of rogue agents raiding accounts, but rather a sign that consumers are letting automated services act for them, including on financial matters. The problem, per Axios, is that most banks cannot tell whether the agent activity they observe was requested by the user or not. Bankers interviewed by the outlet described the situation as "horrifying," and a Chase spokesperson told Axios that customers "should not give their usernames and passwords to anyone."

Why it matters

Credential sharing is currently the default way agents get things done on the web, and it breaks the basic security model of password-based login: once a secret is handed over and cached, the owner loses sight of where it lives and who can use it. 1Password for Agents proposes a different arrangement — the agent proves who it is, the human approves each login, and the secret never leaves the vault.

That model also speaks to the problem the Axios report makes concrete: services on the receiving end, particularly banks, currently have no reliable way to distinguish authorised agent activity from unauthorised access. A verifiable handshake between agent, user, and credential manager is a plausible foundation for that distinction, though whether banks and other services will recognise it remains an open question. For now, the tool gives cautious users a practical way to delegate tasks to AI assistants without handing over the keys.