Trusted Access for Cyber Program Expansion and GPT-5.4-Cyber Launch
Key Points
- GPT-5.4-Cyber model with enhanced cybersecurity capabilities and binary reverse engineering
- Expanded TAC program with multiple access tiers and automated identity verification
- Codex Security fixed 3,000+ critical vulnerabilities since launch
Summary
OpenAI is scaling its Trusted Access for Cyber (TAC) program to thousands of verified defenders and hundreds of teams protecting critical software. The company is introducing GPT-5.4-Cyber, a fine-tuned model variant with enhanced cybersecurity capabilities and reduced refusal boundaries for legitimate defensive work, including binary reverse engineering.
Key Points
- Expanded TAC Program: Multiple access tiers for individual users and enterprises with automated identity verification to reduce friction on cybersecurity tasks
- GPT-5.4-Cyber Model: Purpose-built variant with cyber-permissive training, enabling advanced defensive workflows like binary reverse engineering and vulnerability analysis
- Codex Security Impact: Contributed to fixing over 3,000 critical and high-severity vulnerabilities since recent launch
- Three Core Principles: Democratized access through objective criteria, iterative deployment with continuous improvement, and ecosystem resilience investment
- Scaled Defenses: Cyber-specific safety training integrated across GPT-5.2, GPT-5.3-Codex, and GPT-5.4 releases
- Access Pathways: Individual verification at chatgpt.com/cyber; enterprise teams request through OpenAI representatives
- Dual-Use Risk Management: More permissive models deployed with stricter controls, especially for no-visibility uses like Zero-Data Retention
Implementation
Defenses scale alongside capability improvements. The approach balances broad access to general models with granular controls for higher-risk capabilities, supported by stronger verification and trust signals rather than manual gatekeeping.