The European Union has officially begun enforcement of its landmark AI Act, introducing strict regulations, bans on high-risk AI applications, and hefty fines of up to €35 million or 7% of global revenue, marking a major shift in global AI governance that balances innovation with accountability and ethical oversight. (Source: Image by RR)

As the U.S. and China Build Larger AI Models, Europe Focuses on Trust and Accountability

The European Union has officially begun enforcement of its landmark AI Act, introducing strict regulations and potential fines for non-compliant AI applications, marking a major step in global AI governance. The first phase of the law, effective as of Sunday, imposes outright bans on AI systems deemed to pose “unacceptable risk,” including real-time facial recognition, social scoring, and manipulative AI tools. Companies found violating these regulations could face severe penalties of up to €35 million ($35.8 million) or 7% of global revenue, a higher threshold than the GDPR’s existing privacy fines.

While this enforcement marks a milestone, the AI Act remains a work in progress, with further standards, guidelines, and secondary legislation still in development. According to Tasos Stampelos, Mozilla’s head of EU public policy, the AI Act is primarily a product safety regulation, meaning compliance will evolve over time as additional rules are established. The EU AI Office, as reported in cnbc.com, has already drafted a second code of practice for general-purpose AI models (GPAI), including risk assessments for high-impact AI systems while exempting some open-source models from compliance burdens.

Despite its ambition, the AI Act has drawn criticism from technology executives and investors, who fear that overregulation could stifle innovation. Critics, including Prince Constantijn of the Netherlands, argue that Europe is prioritizing regulation over AI development, potentially hindering its global competitiveness in the industry. Others worry that fast-moving AI advancements make it difficult to create long-term, adaptable legislation that keeps pace with technological progress.

However, some industry leaders believe that the AI Act positions Europe as a global standard-setter in responsible AI. Diyan Bogdanov, an AI expert at Payhawk, argues that Europe’s focus on bias detection, risk assessments, and human oversight defines a new benchmark for AI safety rather than limiting progress. While the U.S. and China dominate in AI scale, proponents of the Act believe that Europe is leading in AI trustworthiness, potentially influencing global AI governance and ethics frameworks in the years ahead.

read more at cnbc.com