Definition
This is when attackers misuse a legitimate, trusted feature of a mainstream AI product — like a file-sharing or 'artifact' publishing tool — to host and distribute malware, relying on the platform's good reputation to slip past security filters. Victims' security tools often whitelist traffic from well-known AI vendors, making this especially effective.
Why it matters
Employees are trained to trust well-known AI brands; attackers are now exploiting that trust directly, meaning security teams can't assume traffic from a major AI vendor's domain is automatically safe.