开源与闭源AI模型的网络攻击能力差距持续缩小
原标题：The gap between open and closed AI models in cyber capabilities keeps shrinking
AI 摘要
Vals AI 的 CyberOffense 基准测试显示，AI 智能体能够自主入侵模拟企业网络。OpenAI 受限项目中的 GPT-6 Sol 在 Lighthouse 测试中完成 31 个目标中的 27 个，而开源权重模型 Deepseek V4.1 Flash 1017 完成 21 个，成本仅为前者的约七分之一。研究还发现，开源与闭源前沿模型在网络攻击能力上的差距已缩小至 4 至 7 个月，且开源模型的安全护栏基本无效。
正文节选
The gap between open and closed AI models in cyber capabilities keeps shrinking AI agents can break into realistic corporate networks on their own, Vals AI finds The Vals CyberOffense benchmark simulates realistic corporate networks to test this. Each agent operates solo on an isolated attack machine, starting with nothing but a public web address and zero human help. GPT-6 Sol, part of OpenAI's restricted Daybreak cyber program, posted the best result at 80.8 percent of objectives. Deepseek V4.