# Companies seek inclusion on White House AI trusted partner list > Source: > Published: 2026-09-11 14:35:12+00:00 # Companies seek inclusion on White House AI trusted partner list A new executive order created a vetting framework for access to frontier AI models, but nobody seems to know the exact rules for getting in The White House’s AI “trusted partner” framework, born from a June 2026 executive order, has created a new gatekeeping layer between frontier AI models and the organizations that want to use them. Companies like OpenAI and Anthropic are already limiting access to their most advanced systems based on government approval, but the criteria for earning that approval remain frustratingly opaque. ## The new gatekeepers The executive order established a voluntary review process for government approval of advanced AI models before they can be released to trusted partners. The framework is designed to assess cybersecurity risks associated with frontier models, the kind of systems powerful enough to reshape digital offense and defense. OpenAI moved quickly in response. The company decided to limit its upcoming model, GPT-5.6 Sol, exclusively to partners who’ve been approved through the new vetting process. Anthropic followed a similar path, restricting access to its Mythos 5 model to a small group of approved cyber defense professionals. Previously, AI developers themselves managed their own access lists. Now, the national security apparatus has inserted itself into that process, and the companies appear willing to go along with it. ## Gold Eagle and the cybersecurity angle In July 2026, the White House launched what it’s calling the Gold Eagle AI cybersecurity clearinghouse. The initiative is meant to consolidate partnerships with critical infrastructure operators, creating a centralized hub for deploying AI-driven cyber defenses at national scale. The problem is that the rules themselves are still a work in progress. As of mid-September 2026, the specific qualification criteria for becoming a trusted partner have not been publicly disclosed. Neither have the reporting standards that approved partners will need to meet. ## A global scramble with fuzzy guidelines The ambiguity isn’t just frustrating domestic companies. Discussions among G7 representatives and US officials about access to frontier AI models took place as early as mid-June 2026, shortly after the executive order was signed. Allied nations find themselves in an awkward position, maneuvering for inclusion or trying to carve out alternative pathways to the same technology. **Disclosure:** This article was edited by Editorial Team. For more information on how we create and review content, see our [Editorial Policy](https://cryptobriefing.com/editorial-policy/).