AI Governance and Compliance
The EU bans practices by risk tier, China works through filings and security assessments, the US leans on executive orders plus the NIST framework — the same product faces entirely different gates in each market.
The EU grades AI systems into four risk tiers, banning social scoring and real-time public facial recognition while demanding transparency and human oversight in hiring and education; China relies on algorithm filing, security assessment and content labeling; the US works through executive orders plus the NIST framework. For cross-border launches, keep a per-market record list: data sources, evaluation logs, labeling and human review.
