EU AI Act Demystified: Risk Tiers, the August 2026 Rules
Details
⚠️ IMPORTANT REGISTRATION NOTICE
(YOU HAVE TO JOIN THROUGH LUMA TO BE REGISTERED FOR THE EVENT (LINK PROVIDED).
EU AI Act Demystified: Risk Tiers, the August 2026 Rules, and What's Still Coming
Format: 90-minute presentation and panel, including live Q&A
Audience: Early-stage to mid-stage AI founders, technical leads, product managers, and compliance advisors
Welcome and intro (15 min)
Part 1: The Four EU AI Risk Tiers (25 min)
Every system placed on the EU market lands in exactly one tier, and the tier, not the technology, determines what's owed.
- Unacceptable Risk, social scoring, subliminal manipulation, emotion recognition in workplaces and schools, among other banned practices. No exceptions, no path to compliance, banned outright.
- High Risk, recruitment, credit scoring, medical devices, educational admissions, critical infrastructure. Two different routes into this tier, not one, and both are currently postponed: standalone systems to 2 December 2027, embedded systems to 2 August 2028.
- Limited Risk, chatbots, synthetic media, deepfakes. Transparency and marking duties, active now, enforceable since 2 August 2026.
- Minimal Risk, spam filters, game NPCs, recommendation engines. No tier-specific obligation, but not zero obligation: Article 4's AI literacy duty reaches this tier too.
Part 2: Active Enforcement vs. Future Milestones (15 min)
- Currently enforced: Article 50 disclosure duties for direct AI interaction; Article 4 AI literacy, in force since February 2025, though worth being precise, it's not directly finable at EU level the way the other tiers are; enforcement runs through national penalties and as an aggravating factor in other findings.
- Postponed: standalone high-risk to December 2027, embedded high-risk to August 2028, national AI regulatory sandboxes to August 2027.
Part 3: The Founder's Toolkit (10 min)
The classification takeaway tool explained.
Part 4: Interactive Q&A (15 min)
An opportunity to ask questions about the EU AI Act.
## About Cinthify
Cinthify exists on a simple belief: governance works when it's proactive, built into the people already making the decisions, not reactive, bolted on by someone external after something's gone wrong. That's the thinking behind everything above, this session, the workshop, and the literacy training that follows it. Cinthify works with founders and technical teams who want that judgment in the room early, not audited after the fact.