Overview
Severity: MEDIUM | Affected: EU AI Industry | Category: policy
The European Union has officially begun enforcement of key provisions within the landmark AI Act. As of this month, companies deploying AI systems classified as 'high-risk' within the EU market are now legally required to undergo rigorous third-party conformity assessments and provide comprehensive documentation on their training data provenance. The high-risk category includes AI used in critical infrastructure, medical devices, employment, and law enforcement. The enforcement, led by newly established national supervisory authorities, means companies must demonstrate robust risk management, high levels of accuracy, and human oversight. Non-compliant organizations face steep fines of up to €35 million or 7% of their global annual turnover. This move represents one of the most significant regulatory actions in AI safety and is expected to set a global precedent for AI governance and accountability, forcing companies to re-evaluate their development and deployment lifecycles.