OpenAI Launches GPT 5.4 Cyber Model to Expand Trusted Access Cybersecurity Program

OpenAI Expands Trusted Access for Cyber Program with GPT 5.4 Cyber Model

OpenAI announced the expansion of its Trusted Access for Cyber program to include "thousands of individuals and organizations," enabling them to use advanced cybersecurity tools to identify and address vulnerabilities in their products. The initiative now incorporates GPT 5.4 Cyber, a new variant of ChatGPT specifically optimized for cybersecurity tasks.

The company’s goal is to make advanced cybersecurity tools more widely accessible while maintaining strict controls to prevent misuse. Access to the program and the cybersecurity-focused model will still be governed by "strong" Know-Your-Customer (KYC) and identity verification rules.

"Our goal is to make these tools as widely available as possible while preventing misuse. We design mechanisms which avoid arbitrarily deciding who gets access for legitimate use and who doesn’t."

Program Expansion and Industry Impact

OpenAI’s announcement follows Anthropic’s recent rollout of Project Glasswing, a similar initiative aimed at providing major tech companies with Claude Mythos, an unreleased model deemed too dangerous for commercial sale. While OpenAI has avoided direct comparisons to Mythos, both models are designed for cybersecurity applications.

Cybersecurity experts in the U.S. and U.K. have described Mythos as a significant improvement over previous frontier models in identifying and potentially exploiting cybersecurity vulnerabilities. However, debates continue regarding its ultimate impact on information security. Similarly, GPT 5.4 Cyber has been fine-tuned for vulnerability research and testing, with OpenAI planning iterative improvements based on lessons learned.

OpenAI’s Approach to Cybersecurity Access

OpenAI has publicly discussed the Trusted Access for Cyber program for months, emphasizing its commitment to enabling broad access while minimizing misuse. The company has stated it does not intend to centrally control which industries or sectors participate in the program.

"We don’t think it’s practical or appropriate to centrally decide who gets to defend themselves. Instead, we aim to enable as many legitimate defenders as possible, with access grounded in verification, trust signals, and accountability."

OpenAI plans to allow a broader group of cyber operators to use the model to protect critical infrastructure, public services, and other digital systems. The company remains cautious about overreach, prioritizing decentralized access over centralized control.

The announcement underscores OpenAI’s focus on balancing accessibility with security, ensuring that advanced cybersecurity tools reach those who can use them responsibly.

The post OpenAI expands Trusted Access for Cyber program with new GPT 5.4 Cyber model appeared first on CyberScoop.