White House Secures Voluntary AI Safety Pledges From Seven Tech Giants
Seven leading AI companies agree to voluntary White House commitments focused on safety, security, and transparency for future generative models. These pledges potentially set the stage for upcoming formal regulations and executive orders.
The White House announces that seven leading AI companies agree to voluntary commitments focusing on the safety, security, and trust of their systems. These companies pledge to conduct rigorous security testing, protect model weights as core intellectual property, and implement watermarking processes for AI-generated audio and visual content to increase transparency.
These forward-looking commitments apply exclusively to the next generation of generative AI models that surpass the current industry frontier. This specific targeting means the agreements do not affect existing widely known models, but instead govern future systems that are more powerful than any currently released technology.
Although the commitments lack formal enforcement mechanisms, the Federal Trade Commission potentially holds the companies accountable through its Section 5 authority against unfair or deceptive practices. The Administration is simultaneously developing an executive order on similar AI governance issues, indicating that these voluntary pledges serve as a stepping stone toward formal regulation.