A high-risk AI system is an AI use case that triggers stricter obligations under the EU AI Act because it can materially affect people, safety, or fundamental rights.
For SaaS teams, the key question is not whether a product "uses AI" in general, but whether the specific use case falls into a regulated category and creates documentation, oversight, testing, incident, and monitoring duties that need named owners.
Related coverage: