1Password launches way for AI agents to log in on your behalf without seeing your password

Concerns are growing about the access that consumers are increasingly entrusting to AI agents. Now, the security services are trying to catch up.

Password manager 1Password launched a new tool Thursday through which users can allow agents to log into their personal accounts on their behalf, without the agent ever having access to a user’s credentials.

The new platform, called 1Password for Agents, achieves this by creating a communication system between the agent, the user and the password manager. When an agent attempts to log in on behalf of a user, a unique digital key is sent to 1Password to confirm the agent’s digital identity. The user then receives an alert about the connection attempt and can personally confirm the request. 1Password then logs in the agent, filling out credentials on behalf of the digital assistant, allowing them to log in and complete any task at hand.

1Password for Agents allows users to choose the agent of their choice, whether it’s Claude Duet from Anthropic, Dots from OpenAI or ChatGPT Work, Grok Bot, Meta’s Muse or Agents from Town, which is an AI assistant startup. The company says it is working with other companies to expand the service.

“Having a password, secret, or ID is a common shortcut to an agent, but can leave people with sensitive information in places they can no longer control because a copy is saved for the next time,” the company said in a blog post. “1Password for Agents provides users with a reliable way to delegate tasks to personal AI agents while maintaining control of the passwords, secrets, and credentials required by those tasks. »

Online agent activity is increasingly disrupting large institutions. Indeed, the new 1Password platform is launched the same day that Axios reports that American banks have started to detect agent activity in user accounts.

This isn’t necessarily about dishonest agents stealing accounts, but rather an example of how consumers are increasingly allowing automated services to act on their behalf, including on financial matters. However, the risk of agent-related issues remains significant. The big problem seems to be that most banks can’t tell whether or not agent activity is related to user requests.

Bankers interviewed by the outlet described the situation as “horrible,” and a Chase spokesperson interviewed by Axios also told the outlet that customers “should not give their usernames and passwords to anyone.”