The AI Act is founded on a risk-based approach. The Draft Guidelines on Classification of High- Risk Systems abandon this approach. Instead, the Draft Guidelines treat any AI system as high risk if other entities in the AI supply chain combine that system with high-risk tools. It also treats general-purpose AI systems as high-risk if any potential downstream high-risk usage is not prevented. These fundamental changes are at odds with the Act and should not be made through secondary guidance. Significant revisions are needed to avoid gutting the Act’s risk - based approach.