HR, Credit, Health: The AI Assets the AI Act Classifies High-Risk

The first question in AI due diligence isn't technical. It's regulatory: does this asset fall under Annex III of Regulation (EU) 2024/1689? If it does, it's classified high-risk — and the full set of obligations under Articles 9 to 15 applies, with the fine exposure that comes with it.

The targeted domains are massive markets

Annex III isn't aimed at marginal cases. It covers sectors where a large share of European AI asset value is concentrated:

• Employment & HR — CV screening, evaluation, hiring or promotion decisions
• Credit & essential services — scoring, access to credit, to insurance
• Education — assessment, admission, examinations
• Biometrics & categorization
• Critical infrastructure, migration, justice, law enforcement

An HR-tech scale-up, a credit-scoring fintech, an assessment edtech: all of them operate potentially high-risk systems. For an acquirer, that's precisely where regulatory risk — and therefore the discount — concentrates.

Classification first, the rest after

Everything starts there. A misclassified asset means either underestimated exposure (and a latent fine) or costly over-compliance. The D7™ score establishes the Art. 6 / Annex III classification first, then characterizes the level of compliance dimension by dimension. It's the answer to the first question every buyer — and every regulator — will ask.