[White House Establishes 'Gold Eagle Program' to Potentially Standardize Review of Initial Users for Cutting-Edge AI Models] According to monitoring by Beating, the White House has established the 'Gold Eagle Program,' responsible for coordinating the discovery of software vulnerabilities in AI, as well as promoting validation, fixes, and patch distribution. CNBC, citing informed sources, reported that this program may also oversee which companies and institutions are allowed early access to cutting-edge AI models. In the future, OpenAI and Anthropic may need explicit government approval when selecting their initial partners. The U.S. government has previously participated in the selection of initial partners for GPT-5.6, and such temporary arrangements may later be incorporated into a unified mechanism. The White House denied to CNBC that the government holds approval authority. The White House stated that company participation in testing remains voluntary, and decisions on when models are released and to whom they are made available ultimately rest with the companies. [Original Link]