AI Security Laws: Anticipating Future Risks [German]

Author: IT BOLTWISE | Source: IT BOLTWISE | Read the full article in German

In a recent report led by Fei-Fei Li, a prominent figure in artificial intelligence, a California working group has urged lawmakers to consider not only current risks associated with AI but also potential future dangers that have yet to be observed. This recommendation comes as discussions around AI regulation intensify, particularly in California, where the need for comprehensive assessments of AI-related risks has become increasingly apparent.

The report, which spans 41 pages, was initiated by Governor Gavin Newsom after he rejected a controversial AI safety bill. It emphasizes the importance of transparency in AI development, suggesting that developers should be required to disclose their safety testing and data acquisition practices. The authors advocate for increased third-party oversight and enhanced protections for whistleblowers to ensure that AI technologies are developed responsibly.

The report proposes a dual approach: trust but verify. This means that while developers should have avenues to address public concerns, they must also submit their testing results for independent review. This strategy aims to bolster transparency in AI development and foster public trust in the technology. The final version of the report is expected to be released in June 2025 and has garnered positive feedback from both supporters and critics of AI policy.

[Read More (translated)]