The European Commission will clarify in upcoming guidelines whether manufacturers of products with embedded AI, governed by EU sectoral safety laws, must comply with the AI Act’s high-risk regime. This follows backlash to a preliminary interpretation in last year’s sector-specific guidance, which suggested that manufacturers using harmonized standards — rather than undergoing third-party conformity assessments — could still be subject to high-risk classification. The commission said on Friday that it has since dropped that guidance document, but it has not ruled out carrying forward the legal interpretation in the forthcoming guidelines and applying it more broadly across all EU product safety laws, as hinted by a recent addition to the Toys Safety Regulation.