BRUSSELS — The Business Software Alliance (BSA) calls on the European Commission to ensure its final Guidelines on the AI Act's high-risk classification rules preserve the Act's risk-based approach, warning that several provisions in the draft guidance could expand the scope of high-risk AI systems beyond what the legislation provides.

"The AI Act established a clear risk-based framework for regulating AI," said Hadrien Valembois, Director, Policy—EMEA, Business Software Alliance. "The Commission's Guidelines should reinforce that framework by providing practical guidance without broadening the scope of high-risk AI systems. That will give companies the certainty they need to innovate while maintaining strong safeguards."

In its submission, BSA urges the Commission to confirm that only genuinely high-risk uses trigger the AI Act's most stringent obligations. BSA also recommends clarifying that AI components should not automatically become high-risk because they are incorporated into larger systems, and that AI systems should qualify as safety components only where they have a direct impact on the health and safety of people or property.