What changed
1Password announced its platform for personal AI agents on October 22. According to the company, a supported agent can ask to sign in, the user approves or declines access, and 1Password fills the credential into the destination rather than giving the password to the agent.
The distinction matters for people who want an assistant to work inside existing accounts. It offers a way to authorize a login without copying a password into a conversation. These are the vendor’s descriptions of the system; CGH has not tested the integrations or independently audited their security.
Availability and limits
The company’s launch blog says the feature carries no additional charge for Individual and Family accounts on 1Password.com. It excludes accounts hosted on 1Password.eu and 1Password.ca at launch. This does not mean the underlying password-manager subscription is free.
Availability varies by partner. The product page describes Muse and Grok Bot integrations as rolling out over the coming days; other connections are still in development. Readers should check the current integration details for their device and account before expecting support.
CGH perspective
Credential protection and permission to act are separate decisions. Even if an assistant never sees a password, a signed-in session can still expose private information or enable account actions. Before approving a login, consider the account’s contents, the task’s scope, and how you will check the result. A protected sign-in is a useful control, but it does not make every subsequent action appropriate.