Global Artificial Intelligence Law

The world's first comprehensive horizontal statutory framework for artificial intelligence. It introduces a four-tier risk classification: Unacceptable Risk (banned practices), High Risk (strict pre-market conformity assessments), Specific Transparency Risk, and General-Purpose AI (GPAI) systemic risk rules.

A landmark executive directive establishing federal safety standards, red-teaming requirements under the Defense Production Act for foundation models exceeding computational thresholds, and active FTC enforcement against unfair AI marketing and automated bias.

The UK utilizes an agile, non-statutory approach empowering existing regulators (ICO, CMA, FCA, MHRA) to apply five cross-sectoral AI principles: safety, security and robustness; appropriate transparency; fairness; accountability and governance; and contestability.

China was the first jurisdiction to impose binding generative AI rules. Providers offering public LLM services must file their algorithms with the CAC, ensure training data truthfulness and copyright adherence, and adhere to socialist core values.

Singapore's globally lauded framework synthesizes nine dimensions of AI governance: accountability, data quality, safety alignment, transparency, incident reporting, testing and assurance, security, content provenance, and human-in-the-loop.