Six vectors for reading the whole field

Competing frameworks target the developers of the largest AI models — mandating safety frameworks, incident reporting to the Attorney General, and civil penalties for noncompliance. Includes creation of ILCompute, a public cloud computing resource.

Extends anti-discrimination protections beyond employment into housing, credit, and public accommodations. Bans surveillance-based dynamic pricing and algorithmic wage discrimination.

Mandates digital watermarking on all AI-generated media, professional disclosure requirements, strict synthetic content labels in political advertising, and consumer notification when interacting with AI systems.

First-of-its-kind regulation of emotionally adaptive AI companions. Requires crisis intervention protocols, persistent machine-disclosure, parental consent for minors, and suicide/self-harm detection.

Institutes default opt-out for AI model training on user data. Creates comprehensive data broker registry. Extends privacy protections to neural and biometric data from consumer devices.

Prohibits public agencies from deploying automated decision systems without meaningful human review. Bans AI in teacher evaluations. Creates state AI governance task force and frameworks for state procurement of AI.