White House Keeps Frontier AI Model Framework Secret

The White House has finalized a framework for evaluating and vetting frontier AI models, but it won't be made public. According to sources, major tech companies like Meta, Nvidia, Microsoft, OpenAI, and Anthropic attended a meeting in Washington D.C. today to review the current draft of the proposal.

The framework is part of an executive order issued on June 2, which mandates its creation within 60 days or by August 1. The directive defines which models are eligible for review and instructs AI labs to submit them to the government prior to public release. However, the process is voluntary, raising questions about how the administration will enforce it.

The decision to keep the framework behind closed doors has been met with criticism from experts. Chris McGuire, Senior Fellow for China and Emerging Technologies at the Council on Foreign Relations, called it 'baffling' and questioned the lack of transparency in regulating powerful AI models.