Andromeda October 2026 Security Incident

On October 13, we identified a security incident involving unauthorized access to credentials in Andromeda’s environment. We disabled the compromised accounts that day and have observed no further unauthorized activity since.

We believe the activity began as part of an industry-wide attack on or about October 9th. It affected a limited number of customer accounts and involved exposure of certain Andromeda and customer credentials. If we believe you were affected, you’ve already heard from us.

Our investigation to date has not detected evidence of unauthorized access to customer models, datasets, or training data within Andromeda’s security perimeter or control.

We have rotated all affected keys and taken steps to harden our infrastructure. We have also engaged third-party cybersecurity teams to support a full forensic investigation. That investigation remains ongoing, and we have not yet established the root cause. Other AI infrastructure providers have reported similar incidents.

We remain in direct contact with our customers and will share additional findings and any further actions as the investigation progresses. Customers with questions or who observe suspicious activity should contact their Andromeda representative.