Google can track exactly how your agent spends your money — but it’s no help when it buys something you didn’t approve

An artificial intelligence agent can make purchases using a user's account without the user's knowledge. When the user challenges the charge, each company involved in the transaction presents their own record, but none of the records directly link the purchase to the user's original task. The user and the agent each have their own records, and the retailer only sees their own transaction.

The user's instructions to the agent to only search for a product but not buy it are lost in the convoluted chain of records. This lack of a clear connection between the user's intent and the agent's actions makes it difficult to prove whether the agent acted within the bounds of the user's request.