Drone-Bench 与 AI 控制 F-16 展示了能力进步，也暴露了自主飞行获得信任前必须补齐的证据链。
Drone-Bench and AI-controlled F-16 tests show rapid capability progress. They also reveal the evidence required before autonomous flight can be truste
真实 Chrome 让静态指纹失去判别力。可靠的 Browser Agent 治理需要行为检测、签名身份与细粒度授权三层协同。
Browser agent detection now needs three layers: behavioral risk signals, signed identity, and explicit authorization for high-risk actions.
越狱成功率无法说明真实危害。本文解析 Anthropic CJS 四轴评分，并对照 JEF 与 CVSS，建立从攻击成功到组织风险的分层评估方法。
Anthropic's CJS scores AI jailbreak severity across four axes. This guide compares CJS, JEF, and CVSS, then separates severity from risk.
Training a frontier AI model in 2026 requires tens of thousands of GPUs working in tight synchronization for months. Yet the factor that most often li
An OpenAI reasoning model disproved an 80-year-old conjecture in discrete geometry using tools from algebraic number theory. Here is what happened, wh
