The AI Act is in force. The absence of fines is structural, not permanent.
📅 July 2025 · The European Artificial Intelligence Regulation is fully legally in effect. However, no formal sanction has been recorded to date. This inaction is neither accidental nor lenient: it results from a regulatory machinery still incomplete and in transition.
📌 Incomplete national authorities
Only 3 of the 27 Member States of the European Union have formalized their supervisory authorities: France (CNIL), Spain (AEPD), and Germany (BfDI).
This means that in 24 countries there is still no institutional basis to open sanctioning procedures. The legal framework is approved. Enforcement capacity is not yet operational.
📌 Lack of key guidelines (art. 6.5)
Article 6.5 of the AI Act introduces the concept of “non‑material influence”, essential for auditing LLMs and generative systems. To date, the final technical guidelines defining how to identify, measure, or limit this influence have not been published.
As long as that ambiguity persists, no national authority has solid technical criteria for imposing sanctions.
📌 Active transition period
The AI Act establishes a transition period until 11 May 2026 for all systems existing before the regulation’s adoption.
During this period: automatic sanctions will not apply; a gradual voluntary adaptation is expected; participation in regulatory sandboxes and frameworks like GPAI is encouraged.
This is not a suspension of compliance, but a window for structural preparation.
✅ Executive note
The lack of fines should not be interpreted as political leeway or a legal vacuum. It is a phase of institutional alignment that will expire in less than a year. Organizations that do not act during this period will be directly exposed to enforcement in 2026.
🔎 Institutional audit · Artefact A1 · Version 1.0 · Date: 2025‑07‑30
Issued by Prospective Analysis Unit · delatorre.ai
Classified as Strategic narrative component for regulatory foresight
Integrated into action framework: AI Act · Monitoring Q3 2025 – Q2 2026
✔️ Validations applied (internal AI)
• IA_STRATEGY.regulatory_impact: ✔ Contextual relevance verified
• IA_COMMUNICATION.executive_level: ✔ Tone and narrative hierarchy
• IA_GOVERNANCE.normative_traceability: ✔ Legal consistency confirmed
Technical identifier: A1-2025-V1 · Full review · Ready for public deployment.
