California's AI Transparency Act forces watermarks, detection tools
AB 853 hits OpenAI, Google, Meta with $5K daily fines
Deep Dive
California's AB 853, the AI Transparency Act, took effect August 2, 2026, requiring major AI providers—including OpenAI, Google, Meta, Midjourney, and xAI—to offer free detection tools and embed invisible watermarks in all generated media. Noncompliant companies face civil penalties of $5,000 per day.
Key Points
- AB 853 took effect August 2, 2026, covering OpenAI, Google, Meta, Midjourney, and xAI
- All covered AI providers must offer free detection tools and embed invisible watermarks in generated media
- Noncompliance carries civil penalties of $5,000 per day per violation
Why It Matters
Verified AI provenance becomes standard in California, reducing deepfake fraud and forcing proactive transparency from major model makers.