精选理由
OpenAI出了个Daybreak Red，能让安全研究员用GPT-5.6-Cyber这类专用模型搞漏洞测试，比通用模型更对口。
OpenAI发布Daybreak Red，一个面向授权安全研究的安全模型访问平台。该平台提供GPT-5.6-Cyber等专用模型，用于漏洞研究、利用验证和安全测试。Daybreak Red面向有经验的安全防御者，支持处理复杂且已获授权的网络安全挑战。
原文 · OpenAI
Daybreak Red provides access to purpose-trained cybersecurity models, including GPT-5.6-Cyber, for a...
Daybreak Red provides access to purpose-trained cybersecurity models, including GPT-5.6-Cyber, for authorized vulnerability research, exploit validation, and security testing. It’s designed for experienced defenders working on complex, authorized cybersecurity challenges. 💬 7 🔄 19 ❤️ 465 👀 40192 📊 48 ⚡