White House Establishes Gold Eagle Plan to Review Early AI Model Access for OpenAI, Anthropic

According to CNBC, the White House has established the Gold Eagle Plan to coordinate software vulnerabilities discovered by AI systems and facilitate verification, patching, and distribution. The initiative may also oversee approvals for which enterprises and institutions gain early access to advanced AI models, potentially requiring OpenAI and Anthropic to obtain government clearance before selecting initial partners.

The White House denied to CNBC that the government holds formal approval authority, stating participation in model testing remains voluntary and companies retain final decisions over release timing and user access.