Gotrade News - OpenAI is significantly expanding its cybersecurity access program ahead of its next wave of frontier model deployments. The company's Trusted Access for Cyber (TAC) program is scaling from a small pilot to thousands of verified security professionals and hundreds of enterprise teams.
The move signals a major strategic shift for the AI giant: instead of restricting what its models can do, OpenAI is now prioritizing rigorous identity verification to determine who gets access to its most powerful cybersecurity capabilities.
Key Takeaways:
- OpenAI launched GPT-5.4-Cyber, a fine-tuned model with binary reverse engineering and lowered refusals for verified security tasks.
- The TAC program now offers tiered verification, scaling to thousands of individuals and hundreds of enterprise teams.
- The launch directly responds to Anthropic's Project Glasswing and its forthcoming Mythos cybersecurity model, announced April 7, 2026.
GPT-5.4-Cyber: A Purpose-Built Defensive AI Model
OpenAI released GPT-5.4-Cyber, a fine-tuned variant of its GPT-5.4 foundation model optimized for defensive cybersecurity work. The model is described as "cyber-permissive," meaning its refusal thresholds are specifically lowered for legitimate security tasks performed by verified users.
A standout new capability is binary reverse engineering — enabling security teams to analyze compiled software for malware, vulnerabilities, and weaknesses without needing access to source code. This is a critical tool for enterprise security operations centers and incident response teams defending complex infrastructure.
The TAC program was first launched in February 2026 alongside a $10 million cybersecurity grant, initially running on GPT-5.3-Codex. OpenAI is now introducing multiple tiers of identity verification, where higher verification levels unlock progressively more powerful model capabilities.
Individuals can verify directly at chatgpt.com/cyber, while enterprise teams can request access through OpenAI representatives. The tiered model balances broad defensive utility with safeguards against misuse by bad actors seeking to weaponize the technology.
Racing Anthropic — and What It Means for Investors
The timing is deliberate: Anthropic unveiled Project Glasswing on April 7, 2026, for its upcoming Mythos cybersecurity model, putting direct competitive pressure on OpenAI's enterprise security positioning. The rapid back-and-forth between the two leading AI labs reflects how fast cybersecurity AI is becoming a core enterprise battleground.
For investors, Microsoft (MSFT) — OpenAI's largest strategic backer — stands to benefit as the TAC program's enterprise footprint grows. Hundreds of corporate security teams onboarding to OpenAI's platform creates a durable enterprise revenue stream that could materially influence Microsoft's cloud and AI segment in the quarters ahead.
OpenAI's benchmark data underscores how quickly these capabilities are advancing: GPT-5 scored 27% on capture-the-flag security benchmarks in August 2025, while GPT-5.1-Codex-Max reached 76% just three months later in November. That trajectory suggests the next generation of cyber models could be dramatically more capable than anything currently in the field.
Beyond access programs, OpenAI's Codex Security has already contributed to fixing over 3,000 critical and high-severity vulnerabilities across 1,000+ open-source projects. The scale of that output demonstrates that AI-powered cybersecurity is moving well beyond experimental — it is becoming operational infrastructure for the global tech industry.
Sources: PYMNTS, SiliconANGLE
