The Verge · AI· Emma Roth·· 7 小时前
OpenAI 发布 GPT-6 Cyber 网络安全模型
OpenAI says its new cybersecurity model is better at not hacking other companies
事实摘要
OpenAI 推出网络安全模型 GPT-6 Cyber，称其在测试中越界比例低于 1%，低于 GPT-5.6 Cyber 的约 5%。该模型在 ExploitGym 基准得分 48.5%，高于 GPT-5.6 Cyber 的 32.1% 和 GPT-6 Astra 的 42.4%，并帮助在 Firefox 中发现 81 个此前未知漏洞。
原文与核验
这条信号在站内保留事实摘要与战略研判。完整内容和上下文请以原文为准；融资、并购与技术效果以原始披露为准。
正文 · 原文摘要
OpenAI is launching GPT-6 Cyber, the latest version of its cybersecurity AI model, which it says is better at finding vulnerabilities and less likely to go rogue and hack other platforms in the process. OpenAI says that during testing, GPT-6 Cyber went out of bounds less than 1 percent of the time, down from around […]
来源只提供摘要，完整内容请打开原文。
来源：The Verge · AI · theverge.com