I Read EU AI Act Article 6 So You Don't Have To — Here's the Risk Checklist

Last month I spent an afternoon reading the EU AI Act — all 144 pages. Most of it is institutional boilerplate. But Article 6 stopped me cold. Article 6 is the one that decides whether your AI system is "high-risk" or not. And if it is, you're looking at mandatory risk assessments, conformity procedures, registration in an EU database, and fines up to €35M if you skip them. The problem? The legal text is dense. Cross-references everywhere. Annex I, Annex III, "harmonised legislation" — none of it maps cleanly to how developers think about their systems. So I built myself a checklist. Five questions, straight from the text, translated into terms a dev can actually answer. What Article 6 Actually Says Here's the short version. Article 6 defines two paths to "high-risk": Path 1 — Annex I (Safety components): Your AI is a safety component of a product already covered by EU harmonised legislation (think: medical devices, machinery, vehicles, toys). If the product requires third-party confor