OpenAI launches GPT-5.6 Cyber model and expands Daybreak cyber defense service to two tiers

OpenAI expanded its Daybreak cyber defense service into two tiers, Blue and Red, both offering access to frontier cyber-focused models under limited access controls. The Red tier introduces GPT-5.6 Cyber, a purpose-trained model built on GPT-5.6 Sol, available initially to named enterprise partners including Accenture, IBM, CrowdStrike, and Cloudflare. The move follows Anthropic's release of its own cyber-focused model Mythos, suggesting a pattern of frontier labs productizing restricted-access vertical models for high-stakes enterprise use cases. For financial institutions, the emergence of tiered, restricted-access cyber AI services signals a nascent vendor category where switching costs and partner vetting requirements may be structurally higher than in general-purpose AI tooling. Also reported by: https://techcrunch.com/2026/08/10/as-ai-led-attacks-multiply-openai-launches-a-new-cyber-model/