📄 中文摘要
当一个 AI 模型以 100% 的置信度将一位国家元首的真实视频标记为深度伪造时，表明数字证据验证架构存在根本危机。最近内塔尼亚胡咖啡馆视频事件中，AI 聊天机器人错误地将真实视频标记为深度伪造，突显了“黑箱”检测在法律和调查领域的风险。对于为私人调查员和开放源情报专业人士构建工具的开发者而言，检测算法的技术含义非常明确：需要改进现有的验证机制，以避免类似错误的发生。
AI 将内塔尼亚胡咖啡馆视频称为深度伪造。实际上并非如此。这才是真正的问题。
📄 English Summary
AI Called Netanyahu's Café Video a Deepfake. It Wasn't. That's the Real Problem.
When an AI model confidently labels a verified video of a head of state as a deepfake, it reveals a fundamental crisis in the architecture of digital evidence verification. The recent incident involving Netanyahu's café video, where an AI chatbot incorrectly identified a real video as a deepfake, highlights the risks of 'black box' detection in legal and investigative sectors. For developers creating tools for private investigators and OSINT professionals, the technical implications are clear: there is an urgent need to improve existing verification mechanisms to prevent such errors from occurring.