InfoQ 研讨会：生产环境 AI 的静默失败、测试工程与智能体治理
原标题：AI in Production: Panellists Highlight Silent Failures, Harness Engineering, and Agent Governance
AI 摘要
InfoQ 举办线上研讨会，ThoughtWorks 的 Zichuan Xiong、Prem Chandrasekaran，隐私专家 Katharine Jarmul 与金融 AI 从业者 Lan Chu 等资深工程实践者讨论了智能体系统在生产环境中的真实运维挑战。与会者指出 LLM 流水线存在语义漂移、嵌入漂移等“静默失败”，RAG 可能泄露敏感文档，GitSpawn 漏洞可借恶意仓库配置执行未授权代码。针对编码智能体产出代码速度超过人工审查的问题，他们建议在 CI/CD 中自动化风格检查、重拾变异测试，并将评估视为持续运行的基础设施。安全方面则主张红队测试、为智能体分配非人类身份、分级访问控制、对 MCP 依赖做版本管理，并强调部署审批最终仍是业务战略决策。
正文节选
During a recent InfoQ webinar titled "AI in Production: What Breaks, What Works, and Who Approves It?", a panel of senior engineering practitioners met to dissect the real-world operational challenges of deploying agentic systems. The discussion was moderated by Hien Luu and featured Zichuan Xiong and Prem Chandrasekaran from ThoughtWorks, privacy specialist Katharine Jarmul, and financial AI practitioner Lan Chu. Luu opened the session by noting three core drivers: AI fails differently from tra