$CCD Google’s Agent Identity is useful because it gives agents a cryptographic identity and clearer authorization inside its cloud environment. But I think the harder problem begins when an agent leaves that environment. A company interacting with an external agent needs more than “this workload is authentic.” It needs to know which real entity is accountable for deploying and operating it. That’s where Concordium’s Agent Registry is interesting. The agent can be linked to an identity-backed account, creating a verifiable accountability trail that isn’t limited to one organization’s internal systems. As agents start doing business across company boundaries, knowing where an agent runs may not be enough. Knowing who stands behind it could become just as important.