OpenAI Expands Cyber Defense with GPT-5 Program GPT-5.4-Cyber

OpenAI has officially launched the expanded phase of its Trusted Access for Cyber program. This initiative provides select organizations with access to its specialized GPT-5.4-Cyber model, aiming to strengthen digital defenses across critical infrastructure, financial services, and open-source security communities.

The program operates on a tiered trust model advanced AI cyber capabilities are made broadly available to defenders, but access scales based on validation, accountability, and demonstrated safeguards.

OpenAI positions this as a direct response to the growing asymmetry between attackers leveraging AI tools and defenders who often lack equivalent resources.

Who Has Joined the Program

OpenAI confirmed that several major enterprises and cybersecurity firms have already signed on.

Including Bank of America, BlackRock, BNY, Citi, Cisco, Cloudflare, CrowdStrike, Goldman Sachs, iVerify, JPMorgan Chase, Morgan Stanley, NVIDIA, Oracle, Palo Alto Networks, SpecterOps, and Zscaler.

These organizations will use GPT-5.4-Cyber to enhance real-world defensive operations, generate threat intelligence, and help OpenAI refine safety systems through practical deployment feedback.

OpenAI also granted access to the U.S. Center for AI Standards and Innovation (CAISI) and the UK AI Security Institute for independent testing, reinforcing its commitment to third-party oversight.

Recognizing that most software teams lack 24/7 security operations coverage, OpenAI committed $10 million in API credits through its Cybersecurity Grant Program to extend access to its frontier models to under-resourced defenders.

Initial grant recipients include:

• Socket and Semgrep: focused on software supply chain security.
• Calif and Trail of Bits:  pairing AI with expert vulnerability researchers.

OpenAI emphasized the real-world problem this addresses; not every team can respond to a critical vulnerability disclosed on a Friday night.

The grant program aims to change that by giving smaller open-source maintainers and researchers the same AI capabilities available to large enterprises.

Additional teams with proven track records in open-source and critical infrastructure security can apply directly through OpenAI’s grant portal.

The Defense-First Philosophy

OpenAI’s framing is clear: cyber defense is a collective challenge. The program is designed to generate shared learnings across participants, improve model safety through real-world use, and push the frontier of defensive research.

BNY’s Chief Information Officer, Leigh-Ann Russell, noted that the firm’s participation reflects its commitment to protecting financial system resilience as AI capabilities accelerate, building on an existing collaboration with OpenAI.

The company confirmed that Trusted Access for Cyber will continue to expand, with safeguards that increase in step with model capability, ensuring that greater power comes with proportionally stronger accountability measures.

Disclaimer: HackersRadar reports on cybersecurity threats and incidents for informational and awareness purposes only. We do not engage in hacking activities, data exfiltration, or the hosting or distribution of stolen or leaked information. All content is based on publicly available sources.