OpenAI发布网络安全专用模型GPT-5.6-Cyber
Expanding Daybreak as the Cyber Defense Window Narrows
OpenAI发布GPT-5.6-Cyber，能用Daybreak Red做漏洞研究。
OpenAI推出网络安全专用模型GPT-5.6-Cyber。该模型通过Daybreak Red向授权用户开放。GPT-5.6-Cyber可用于漏洞研究、漏洞验证与安全测试。
Expanding Daybreak as the Cyber Defense Window Narrows
Meet GPT-5.6-Cyber, OpenAI’s cybersecurity-specific model available through Daybreak Red for authorized vulnerability research, exploit validation, and security testing.
This report is categorized under AITOP's "Models" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include GPT-5.6-Cyber, Daybreak Red, OpenAI, AI安全.
Primary source verification: published directly by OpenAI Blog with first-party authority. Cross-source corroboration: 11 independent sources have covered this development (including Greg Brockman, Decoder, Sam Altman), confirming industry-wide signal strength.
