We’re planning to add AI agents as collaborators on merchant stores to help with store issues and app configuration. The blocker: collaborator requests require identity verification, and that flow assumes a human on the other end with a real name, ID, and the usual documents.
An agent doesn’t have any of that. So how are people getting agents through an identity verification step that was built for humans?
For those of you who put LLMs and new tooling into production early, I’m guessing some of you have already hit this wall. What worked? Do you verify a responsible human and run the agent under that verified identity? Register the agent as its own entity somehow? Something else entirely?
Would appreciate hearing how you’ve approached it before we commit to an implementation.