思科为 AI Defense 引入 Claude Mythos 做智能体红队测试：同步升级 Splunk 智能体可观测与 Cisco IQ
美国科技公司思科（Cisco）11月9日在澳大利亚 Melbourne 举行的 Cisco Live 亚太、日本及中国区大会上宣布，其 AI Defense 自适应超级智能（SI）红队框架将引入 Anthropic 的 Claude Mythos，利用前沿对抗推理模拟对智能体（AI agents）的真实攻击，帮助防御方在系统上线前做压力测试。
思科同时发布 Splunk 智能体可观测（Agent Observability）与 Cisco IQ 的升级：Cisco IQ 依托数十年的漏洞与故障排查数据，现可把实时威胁检测数据与可能暴露的网络基础设施以机器速度关联，加速修复；其强化的 Tokenomics 功能把智能体的 token 支出与投资回报实时连接，帮助机构优化超级智能（SI）投入。
思科总裁兼首席产品官 Jeetu Patel 表示，无法被保护、治理和观测的东西就无法规模化；思科把前沿级运行时保护与机器速度的智能体质量控制结合，构建一个持续的信任闭环：既保护智能体免受侵害，也确保每一笔 token 支出都产生可衡量的业务价值。
思科引用 Splunk 上月发布的全球调研称，81%的 IT 运维与工程专业人士表示安全与可观测团队之间的协作正在加强。思科称，自2025年1月推出以来，AI Defense 已帮助数百家客户在超级智能（SI）模型、应用与智能体进入生产前发现关键缺陷。
（来源：思科）
Cisco Unveils Frontier AI Red Teaming and Agent Observability to Secure the Agentic Enterprise
Cisco (NASDAQ: CSCO) announced on Nov. 9 at Cisco Live APJC in Melbourne, Australia, that it is adding Anthropic's Claude Mythos to AI Defense for red teaming AI applications and agents, empowering defenders to stress-test AI systems with frontier adversary reasoning.
Coupled with new Splunk Agent Observability and Cisco IQ advancements, Cisco said it is preparing enterprises for a new era of trust that requires securing, observing, and governing agents at machine speed. Cisco IQ, which leverages decades of vulnerability and troubleshooting data, now correlates live threat detection data with potentially exposed network infrastructure at machine speed to accelerate remediation, and enhanced Tokenomics connects agent cost directly with ROI in real time.
"You cannot scale what you cannot secure, govern, or observe," said Cisco President and Chief Product Officer Jeetu Patel. "By uniting frontier-level AI runtime protections with machine-speed agent quality controls, we are building a continuous trust loop: protecting agents from compromise while ensuring every token spent delivers measurable business value."
Cisco cited a global Splunk research report from last month in which 81% of ITOps and engineering professionals cited increased cooperation between security and observability teams. Since Cisco launched AI Defense in January 2025, it has helped hundreds of customers identify critical flaws in AI models, applications, and agents before they reach production; its adaptive red teaming allows security teams to quickly build customized, multi-step attacks on target systems to uncover specific risks, and Claude Mythos will augment this framework with advanced reasoning.
(Source: Cisco)
来源：思科
发布：GSM·美国（发布链接）
图片来源：图：位于 San Jose 的思科总部园区。CC0，经 Wikimedia Commons。 / Photo: The Cisco main campus in San Jose. CC0, via Wikimedia Commons.
GSM Global · 本篇英文版 English edition of this story
