An AI asset doesn't lose its value all at once — it loses it on one of the AI Act's 7 dimensions, and it's the one you didn't cover that the buyer finds. MB AI scores all seven, surfaces your gaps, delivers the plan. Then MB Shield™ defends your level as the law and your product change. Self-serve, from public data to your dataroom. High value, accessible price.
In 2026, one in five buyers walked away from a deal because of AI (Bain). When that risk is the EU AI Act, it attaches to the asset and survives closing — and it "could affect the valuation", market analyses confirm. → Sourced analysis note (tier-1 sources)
Regulation (EU) 2024/1689 has been in force since 1 August 2024. Prohibited practices (Art. 5) apply since February 2025, GPAI obligations since August 2025. The Digital Omnibus deferred compliance for high-risk Annex III systems to 2 December 2027 — risk management, technical documentation, human oversight, EU registration. The regime is not loosening: it is consolidating, and the AI Office's role is reinforced.
Regulatory risk attaches to the asset, not to the transaction. The buyer inherits the obligations and potential fines from closing. No law firm quantifies them. No leading audit firm certifies them in under 6 months.
Start by knowing whether you're in scope. Move up to the full score when you need it. Step up to signed proof for closing. Each step stands on its own — and sets up the next.
Are you in scope of the AI Act? Art.6 classification, Art.71 fine exposure, 3 priority actions. In 24h.
0–100 score across 7 dimensions, from your data. Phased compliance plan included. GO/NO-GO decision before LOI.
D7™ score driven by your dataroom under NDA + a formal signed opinion. The M&A instrument: generated clauses (escrow, R&W, earn-out), defensible at signing, board and regulatory filing.
AI Act compliance isn't a report, it's a state to maintain. Periodic re-scan, targeted alerts, progression plan and re-attestation — your level is defended as the law evolves and your product changes. MB Shield is entering beta: sign up, you may get to take part in the beta.
You name the asset, the sector, the transaction context, the indicative valuation. Optional: dataroom access under NDA.
Public sources in parallel: website, press, GitHub, INPI, CNIL — and your dataroom where applicable. Each source receives a PCAOB/ISA 500 reliability score.
7 weighted dimensions, proprietary D7™ methodology. Global score 0–100, RAG status and compliance plan. Every claim documented and justified.
Per-dimension score, Monte Carlo valuation, phased recommendations 0–30d / 90d / 12 months, contractual clauses (escrow, earn-out, R&W), Opinion status.
If the number of verifiable sources is insufficient, the memo is automatically marked DRAFT — MB AI legal protection, immediate signal for the W&I underwriter.
D7™ is MB AI's proprietary analytical framework for characterizing EU AI Act exposure and the value of an AI asset across seven distinct dimensions. No two assets ever produce the same profile.
The D7™ methodology is a proprietary synthesis of international best practice in AI auditing, applied selectively by dimension to produce the most defensible, evidence-grounded score.
D7™ synthesizes regulatory, technical, commercial and financial risk into a single score — weighted by a proprietary methodology defensible in M&A, W&I insurance and regulatory filings.
Live demo of the MB AI Score™ engine. Enter the asset and the transaction context — get the 7-dimension scoring, the valuation range and the Opinion Letter status. Access restricted to authorized users.
Fill in the asset details
and click Generate to produce
a MB AI Score™
The European AI Office has opened its first GPAI investigations. National authorities — including the CNIL in France — are designated. Every Annex III AI asset without a D7™ score is an unpriced risk in your portfolio.
Choose your format. Name the asset. MB AI delivers the score.
A monthly D7™ extract on a real asset + an EU AI Act intelligence brief. See what the score produces before you order.
Your first D7™ score in 24h: Art.6 classification, estimated Art.71 exposure, 3 priority risks. Know where you stand before any decision — then defended by MB Shield™.
0–100 score across 7 dimensions, from your data. Phased compliance plan included. In 48–72h.
Score on your dataroom (NDA) with a full compliance plan, contractual clauses and a formal signed opinion for closing. Full remediation plan optional.
→ The 12 "AI Act value gap" briefs — tier-1 sources (consulting, M&A, audit)…
Choose your format. Name the asset. MB AI runs the D7™. You receive a defensible memo.
Request a memo →