A systematic 4-Lens Framework that checks your compliance documentation against the Act's requirements — and produces a fully revised version with every gap fixed.
August 2, 2026: High-risk AI system obligations become enforceable. Non-compliance fines: up to €35 million or 7% of worldwide annual turnover.
The Problem
Most AI Act documentation fails for the same structural reasons — not because the intent is wrong, but because the language doesn't hold up under scrutiny.
"Appropriate measures will be taken" — appropriate how? By whom? By when? A market surveillance authority will ask. Your document should already answer.
Annex IV requires 9 mandatory sections in your technical documentation. Article 17 requires 13 aspects in your quality management system. Most organizations miss at least 3.
"Human oversight is ensured" — but can the designated person actually override the system? If oversight is nominal, it's non-compliant under Article 14.
If you've classified a system as not high-risk to avoid obligations, is your Article 6(3) reasoning documented and defensible? This is where fines start.
The Method
Every compliance document is audited through four lenses. Each targets a different failure mode that regulators look for.
Are all required elements present? Correct risk classification? Correct role identification (provider vs deployer)? Annex IV completeness? Article 17 coverage?
Article 3 term consistency. Precision of obligations. Specificity of commitments. Every vague "where feasible" or "as appropriate" flagged.
Can the procedures actually be implemented? WHO does WHAT by WHEN, verified HOW? Are incident reporting timelines realistic (15-day deadline)?
What would a hostile auditor find? Where are the loopholes? Could someone claim compliance while not actually being compliant?
Coverage
The framework checks documentation against specific articles, annexes, and requirements of Regulation (EU) 2024/1689.
How It Works
Paste your compliance document into any LLM you already use — ChatGPT, Claude, Gemini, or your organization's approved AI tool.
Scans your document through all four lenses. Output: article-by-article gap log, severity-ranked findings, compliance scorecard.
Simulates a hostile enforcement audit. Finds every loophole, penalty exposure, and vulnerability a market surveillance authority would find.
Produces a complete revised document with every gap addressed. Includes a change log mapped to specific article references and a before/after scorecard.
Pricing
Every tier includes a commercial license for client work. No per-project fees. No subscription. Self-service — all instructions included.
The core audit framework. Everything you need to run a compliance documentation audit.
The complete professional package. For consultants and compliance teams auditing multiple document types.
Everything in System, shared across your team. For consulting firms, compliance departments, and legal teams.
FAQ
Also Available
The same 4-Lens Framework, applied to any policy document — AI governance, procurement, HR, compliance, or regulatory drafts. For professionals who audit policy documents beyond the EU AI Act.
The general-purpose policy QA system. Same methodology you trust from the EU AI Act pack, applied to all your other documents.
Toolkit from $199 · System from $499 · Agency from $1,499
EU AI Act buyers get early access and a loyalty discount.