EU’s AI Act Delay: Extra Time to Build Real AI Governance, Not Quick Fixes

The European Union’s landmark Artificial Intelligence Act is getting a slightly longer runway for some obligations, and policy watchers say companies should use the time to build lasting AI governance instead of scrambling for last‑minute fixes.

The AI Act is a comprehensive law that ranks AI systems by risk level and imposes strict obligations on high‑risk uses such as biometric identification and critical infrastructure. Recent guidance confirms that EU member states will have more time—until 2027—to stand up “regulatory sandboxes,” controlled environments where companies and regulators can test AI systems together. At the same time, the grace period for transparency obligations on AI‑generated content, like watermarks or clear labeling of synthetic media, has been shortened so that providers must comply by December 2, 2026. The shift reflects growing concern over deepfakes and other deceptive content.

Experts argue that organizations that treated GDPR seriously are already ahead: they understand impact assessments, documentation and internal approval flows, all of which map well onto the AI Act. Companies that rolled out generative AI tools informally, however, will need to quickly define standards for model selection, risk assessments and human oversight. Commentators stress that this shouldn’t be a one‑off “compliance project” but part of a broader, board‑level strategy for responsible AI governance across the business.