OpenAI has been briefing federal agencies, state governments, and Five Eyes allies on a new cybersecurity product over the past week. The firm held an event in Washington, D.C., on Tuesday for approximately 50 cyber defense practitioners to demonstrate the capabilities of its GPT-5.4-Cyber model. Government applicants are undergoing the same vetting process as commercial customers seeking access to its Trusted Access for Cyber program.
Companies and agencies are eager to acquire the latest AI tools, which promise significant advantages for defenders but also pose potential risks if misused. The advanced capabilities could lead to major gains in cyber defense while simultaneously enabling frightening advances for malicious hackers.
The model was rolled out under a tiered access program last week. Attendees at the D.C. event included a range of officials from across the government and national security agencies, most of whom oversee day-to-day cyber tasks.
OpenAI's rollout follows closely behind Anthropic's release of its own model, Mythos Preview. Both companies are now working with government agencies to determine which entities will gain access to these powerful tools.
The briefings highlight the growing intersection of artificial intelligence and national security, where cutting-edge technology is being rapidly integrated into defense strategies.