The European Union (EU) began enforcing its ground-breaking Artificial Intelligence (AI) law on Sunday, which includes stringent rules and heavy fines for violators. The rule, which took effect in August 2024, prohibits AI systems that pose “unacceptable risks,” such as social grading, real-time facial recognition, and manipulative tools, CNBC reported.
Companies now face fines of up to EUR 35 million (USD 35.8 million), or 7% of global annual revenue, whichever is greater.
The AI Act targets applications that categorize individuals by race, sexual orientation, or other sensitive attributes, aiming to protect citizens from harmful uses of AI. The penalties are higher than those that are applicable under the EU’s General Data Protection Regulation (GDPR), which caps fines at 20 million euros or 4% of annual turnover.
While the law is now partially in effect, full implementation will unfold gradually. Tasos Stampelos, head of EU public policy at Mozilla, noted that the AI Act is “product safety legislation” requiring ongoing updates to standards and guidelines. Compliance will depend on secondary regulations and risk assessments, particularly for general-purpose AI (GPAI) models such as OpenAI’s GPT.
The EU AI Office, established to oversee compliance, released a draft code of practice for GPAI models in December. It exempts some open-source models but mandates rigorous risk assessments for “systemic” AI systems.
Critics argue the law could stifle innovation. Prince Constantijn of the Netherlands expressed concern that Europe’s focus on regulation might limit its competitive edge in AI development. However, supporters believe the EU is setting a global standard for trustworthy AI.
Diyan Bogdanov of Payhawk applauded the Act’s focus on bias detection, risk assessments, and human oversight, calling it “leadership” in defining ethical AI.
The EU’s AI Act is a significant step in regulating emerging technologies, balancing innovation with safeguards for citizens. As enforcement begins, the global tech industry watches closely to see how the rules reshape AI development and deployment.