Technology giants Meta and Anthropic, along with Google and OpenAI, were invited to the White House on Tuesday, August 3rd, 2026, to discuss a voluntary framework aimed at enhancing AI security. Under this framework, America’s leading artificial intelligence (AI) developers could grant the US government early access to their most advanced AI models for cybersecurity testing, specifically to assess their hacking vulnerabilities.

Reuters reported that the Trump administration finalized the details of these voluntary cybersecurity tests designed to measure the hacking capabilities of the most advanced American AI models. The administration planned to review the framework with the AI industry during the week of August 3rd, marking a cautious shift from its previously hands-off approach to AI regulation.

According to The Information website, the White House meeting with AI companies, including Google and OpenAI, was to review new controls before their rollout. If the framework aligns with the Trump administration’s June executive order on AI, participating companies could provide designated "covered frontier models" to the federal government for testing up to 30 days before granting selected trusted partners access.

It remains unclear whether the details of this voluntary framework will be made public. The White House, Meta, Google, OpenAI, and Anthropic did not immediately respond to requests for comment.

Sources