Five months. That’s what separates energy executives from the EU AI Act’s high-risk compliance deadline, and Baker Botts’s March 16 analysis makes clear the clock is running. The Act’s most demanding obligations, documentation, conformity assessment, human oversight requirements, take effect on August 2, 2026.
Energy companies are a focus because of how the Act is structured. Legal analysts indicate that AI systems used as safety components in critical infrastructure may fall under Annex III, Section 2 of the Act, which covers electricity, gas, and heating networks. The key threshold: does the AI system operate as a safety component where failure could cause physical harm? If yes, high-risk classification likely applies.
The penalty exposure is significant. According to legal analysis of the EU AI Act, non-compliance penalties for high-risk systems may reach €15 million or 3% of global annual turnover, whichever is higher. That framing matters for global energy companies, 3% of global revenue is a much larger number than a fixed fine.
This connects to a broader enforcement reality: EU member states are still establishing their national supervisory authority infrastructure. Companies watching for enforcement signals should monitor that landscape alongside their own compliance timelines. Baker McKenzie’s EU AI Act resource hub tracks the regulatory architecture as it develops.
The compliance gap for high-risk systems is narrowing. Documentation requirements, conformity assessment procedures, and human oversight mechanisms all need to be in place before August 2 – not on August 2.