By Pekka Valo, Aliant Finland. The European Union is positioning itself as a global leader in regulating artificial intelligence, most notably through the AI Act, which will take full effect by 2026. While this framework aims to ensure safety, transparency, and accountability, it also raises a series of legal and practical challenges.
The AI Act classifies AI systems by risk level, imposing strict requirements on “high-risk” uses in areas such as healthcare, policing, and employment. Critics warn these rules could stifle innovation, particularly for startups and small businesses, making Europe less competitive compared to the U.S. and China.
Although the law is EU-wide, enforcement varies by Member State, with differing capacities and interpretations of “high-risk” systems. These risks creating a patchwork of compliance standards across Europe.
Determining who is responsible when AI causes harm—the developer, deployer, or data provider—remains unclear. The withdrawal of the proposed AI Liability Directive in 2025 left victims without strong legal remedies, and no clear framework exists for civil or criminal accountability.
The AI Act overlaps with GDPR, but issues of bias, profiling, and algorithmic opacity remain unresolved. Ensuring fairness and explainability in automated decision-making continues to be a legal and ethical challenge.
AI systems often rely on copyrighted material for training. While text and data mining are permitted under certain EU rules, creators argue protections are inadequate. Providers must disclose summaries of training data, but enforcement is uncertain.
The Act bans certain intrusive practices—such as social scoring, real-time facial recognition in public, and mass biometric scraping—yet rights groups argue loopholes remain. Concerns persist about the balance between security and civil liberties.
The EU’s AI regulations set an ambitious standard, but they face legal uncertainty, enforcement challenges, and industry resistance. Unless gaps on liability, innovation balance, and fundamental rights are resolved, Europe risks both overregulation and underprotection in the age of artificial intelligence.