The EU Artificial Intelligence Regulation (AI Act) is the comprehensive legal framework governing the development and use of AI systems. It entered into force in 2024-2025.
Risk Levels
- Unacceptable risk: Social credit system, biometric surveillance
- High risk: Health, education, employment, critical infrastructure
- Limited risk: Chatbot, deepfake (transparency). obligation)
- Minimal risk (minimal risk): Spam filter, game AI
Obligations for High Risk AI
- Risk assessment
- Data quality
- Transparency and documentation
- Human audit
- Precision and robustness tests
- CE mark
Implications for Turkish Companies
- All companies offering AI products/services to the EU market are covered
- Obligation to appoint an EU representative
- Extraterritorial impact (GDPR-like)
Penalties
- Prohibited AI: €35 million or 7% turnover
- High risk violation: €15 million or 3% turnover
- Misrepresentation of information: €7.5 million or 1% turnover
Harmonization in Türkiye
- Turkey has not yet enacted a special AI law
- Artificial Intelligence Strategy Document is available
- First regulations are expected from 2025
Practical Recommendations
- Take an AI system inventory
- Risk determine your level
- If there is high risk, an immediate harmonization project
- If there is an EU market, priority
Artificial intelligence and technology law lawyer is recommended.