Anthropic responds to Newsom AI Working Group — emphasizes transparency and security practices

AI Impact Summary

Anthropic supports the California AI Working Group's emphasis on transparency as a key component of responsible AI policy. The group’s focus on disclosure regarding model security, theft prevention, and national security risk testing aligns with Anthropic’s existing practices, including its Responsible Scaling Policy and public documentation of safety testing results. Given the anticipated rapid advancement of AI capabilities – potentially by the end of 2026 – the need for proactive transparency and standardized safety protocols is paramount to fostering trust and mitigating potential risks.