OpenAI Launches GPT-5.4-Cyber Shortly After Rival's AI Model Reveal

• Anthropic's Mythos model has already identified 'thousands' of major vulnerabilities in operating systems, web browsers, and other software through its controlled deployment program
• GPT-5.4-Cyber features a more permissive design with fewer restrictions on sensitive cybersecurity tasks like vulnerability research and analysis, requiring careful vetting of users
• OpenAI is expanding its Trusted Access for Cyber program (launched in February) with new tiers, where higher verification levels unlock more powerful capabilities, with top-tier users gaining access to GPT-5.4-Cyber

OpenAI Launches GPT-5.4-Cyber Cybersecurity Model Following Anthropic's Rival Release

OpenAI unveiled GPT-5.4-Cyber on April 14, a specialized variant of its flagship model designed for defensive cybersecurity operations. The launch follows closely on the heels of competitor Anthropic's announcement of its frontier AI model, Mythos, on April 7.

Key Developments:

Anthropic's Mythos, deployed through the "Claude Mythos Preview" program to select organizations, has already discovered "thousands" of major vulnerabilities across operating systems, web browsers, and other software platforms.

OpenAI's GPT-5.4-Cyber will be distributed through a controlled rollout to vetted security vendors, organizations, and researchers due to its more permissive design capabilities. The model features fewer restrictions on sensitive cybersecurity tasks, including vulnerability research and analysis.

Program Expansion:

The company is significantly expanding its Trusted Access for Cyber (TAC) program, originally launched in February, to include thousands of verified individual defenders and hundreds of security teams protecting critical software infrastructure. OpenAI is introducing new tiered verification levels, with higher tiers unlocking more powerful capabilities. Only users approved for the highest tier will gain access to GPT-5.4-Cyber's full functionality.

Market Implications:

The rapid-fire launches from both AI leaders signal intensifying competition in the cybersecurity AI sector. The controlled deployment approach by both companies reflects growing regulatory and safety concerns around powerful AI models capable of identifying software vulnerabilities. This development positions AI as an increasingly critical tool in defensive cybersecurity, potentially reshaping how organizations approach threat detection and software security assessments.