California Enforces AI Transparency Act: Major Providers Face New Regulations

California's new AI Transparency Act, SB 942, goes into effect on August 2, imposing stringent requirements on generative-AI providers with over one million California users. The law mandates these companies to embed C2PA-compatible provenance in images, videos, and audio, offer a free public detection tool, and allow users to add visible AI labels.

New Regulations for AI Providers

The new regulations come with significant penalties. Providers found in violation of the act face fines of up to $5,000 per day per instance. This move is aimed at increasing transparency and accountability in the use of AI-generated content.

Apple Caps Bug Report Submissions

In related news, Apple has introduced a submission cap and a 30-day cool-off period for Feedback Assistant bug reports. The company cites an overwhelming volume of AI-generated submissions as the reason for this change. Researchers can request higher quotas if needed, but the move reflects a broader trend of managing the influx of AI-generated content.

AI-Related Security Concerns Emerge

Researchers have also demonstrated that AI-assisted code can tamper with data from computerized scans of physical DNA evidence. This finding, reported by the Wall Street Journal, exposes vulnerabilities in the chain-of-custody assumptions that underpin roughly 30 years of forensic casework. The implications are significant for the legal and scientific communities, highlighting the need for robust security measures in AI applications.

Academic Shifts in AI Safety

Fields Medalist Jacob Tsimerman is taking leave from the University of Toronto to join OpenAI, focusing on AI safety. Tsimerman, who recently proved the André-Oort conjecture, argues that mathematicians must engage with AI now to shape its future. His move underscores the growing importance of AI safety in academic and industry circles.

White House Misses AI Framework Deadline

The federal government has missed the August 1 deadline set by Executive Order 14409 for establishing a frontier AI framework. The absence of Federal Register notices, NIST or CISA publications, and OSTP statements on the promised benchmarking process, voluntary disclosure framework, and federal cyber-workforce plans raises concerns about the pace of AI regulation at the national level.

References

• AI News Today, August 2 — Top AI Stories & Live Updates | AI Weekly