Woofun AI reports that the White House has established the "Gold Eagle Initiative" to coordinate the identification of AI software vulnerabilities and facilitate validation, remediation, and patch distribution. Sources indicate the initiative may also oversee which entities receive early access to advanced AI models, potentially requiring explicit government approval for initial partners of firms like OpenAI and Anthropic.

The White House denied having final approval authority, stating that participation in testing remains voluntary. Decisions regarding model release timing and availability ultimately remain with the companies, though previous ad hoc arrangements for GPT-5.6 partner selection may be integrated into this unified framework.