OpenAI provided $10M in API grants to security firms using the new GPT-5.4-Cyber model. This Trusted Access for Cyber program targets systemic vulnerabilities through specialized model access. Participating enterprises will deploy these tools to automate threat detection. The move shifts OpenAI from general-purpose utility toward highly specialized, domain-specific security applications for practitioners.