Cloudflare 用 AI 测试框架探测并加固其 WAF
原标题：Cloudflare Uses an AI Harness to Probe and Harden Its WAF
AI 摘要
Cloudflare 将前沿 AI 模型置于受控测试框架中探测其 Web 应用防火墙（WAF），以已被拦截的攻击载荷为起点，让模型生成并迭代新的变体。在 45 个场景中系统共生成 1,107 次尝试，经人工分诊后留下 49 项发现，其中 48 项涉及命令注入或 SSRF。该工作最终促成 Cloudflare 托管规则集的三项变更：新增 SSRF - Obfuscated Host 和 SSRF - Restricted Protocol 两条检测规则，并改进现有 SSRF - Cloud 规则。
正文节选
Cloudflare placed frontier AI models inside a controlled testing harness to probe its Web Application Firewall (WAF), using blocked attacks as starting points for models to generate and refine new variations. Across 45 scenarios, the system generated 1,107 attempts and left 49 findings after human triage. The exercise ultimately contributed to three changes in Cloudflare’s Managed Ruleset. The experiment began with attack payloads that the WAF had already blocked. Rather than repeatedly replayin