印度CERT-In用Anthropic未发布的Mythos模型查找软件漏洞
印度政府消息人士向CNBC-TV18透露,印度计算机应急响应小组(CERT-In)已获得Anthropic尚未发布的Mythos模型的使用权限,正用它查找漏洞并提出修补建议。不过,模型通过云端提供,这引发了对敏感数据的担忧。

印度政府消息人士向CNBC-TV18透露,印度网络安全机构印度计算机应急响应小组已获准使用Anthropic尚未发布的人工智能模型Mythos。消息人士称,CERT-In正用该模型查找脆弱系统中的弱点,并提出修补建议。
CNBC-TV18称,这一权限让CERT-In拿到了一款在发现网络安全漏洞方面能力尤为突出的工具。这也意味着,有关潜在关键系统的敏感信息可能需要发送到一个云平台上。报道特别指出了这一顾虑。
云端访问引发数据担忧
报道称,该模型以云端方式提供使用,这在政府内部引起了担忧。官员们对把敏感信息发送到云平台心存戒备,因为共享潜在关键系统的数据本身带有风险。
对一个国家级计算机应急响应小组来说,数据问题是实际操作层面的。CERT-In检查的系统可能是关键系统,它分享给云平台的信息可能暴露弱点。YourStory报道称,安全团队日常要处理机密源代码、网络日志和事件报告。把这类信息发送到外部平台,会在隐私、访问权限和监管合规方面引起担忧。
CERT-In并非只依赖Mythos。消息人士告诉该媒体,该机构还在使用另外几款AI模型检测系统漏洞,这些模型的能力大约相当于Mythos的80%至85%。同时使用多款模型,可以帮助该机构找出安全缺口,同时减少外界对敏感信息处理方式的担忧。Mythos的云端部署仍是核心考量。
通过Project Glasswing获得的有限权限
据CNBC-TV18报道,Mythos尚未发布,Anthropic通过Project Glasswing向部分选定机构和政府部门提供了有限的使用权限。该计划聚焦于发现并修复关键信息技术系统中的弱点,目的是帮助参与方在恶意行为者利用之前找出安全缺陷。该模型并未公开提供。
80%至85%这个数字意味着,其他模型覆盖了Mythos的大部分能力,但不是全部。如果Mythos能发现其他模型漏掉的缺陷,这一差距就很重要。YourStory报道称,Mistral AI已公开其Large 4网络安全模型的权重,机构可以下载并在自己的基础设施内运行。这种做法让人对AI在哪里运行、数据如何处理有更大的控制权。但这一方式仍需技术能力和安全流程才能有效管理。
对CERT-In来说,眼下任务很明确:用Mythos查找弱点并提出修复建议,同时权衡它所检查的系统能否接受云端访问。
资料来源
28- 01Govt gets access to Anthropic's most powerful Mythos AI cyber model that can detect vulnerable systems: ReportEN
- 02India's CERT-In gets access to Anthropic's unreleased Mythos AI model for cyber defenceEN
- 03Mistral AI opens trillion-parameter model for cybersecurityEN
- 04Was I wrong about open models & cyber risks?EN
- 05Google reveals new Gemini 4 Carbon AI model, narrowly beating Claude Opus 5.5EN
- 06ChatGPT can now turn text prompts and photos into animated GIFs with OpenAI's Images 2.6 modelEN
- 07Anthropic's text watermark is easy to bypass on several older Claude modelsEN
- 08“War” Secretary Pete Hegseth Endorses 15 Minute City ModelEN
- 09How big is the open model economy today?EN
- 10GPT-6 Astra sees the matrix and reprograms other AI models directly through their weightsEN
- 11Anthropic's new prototype proves which model wrote an AI response without exposing its weightsEN
- 12LiberAI Unveils Liber-0, a World Action Model That Grounds Human Experience in Robot ActionEN
- 13HyperOS 4.0: Xiaomi verteilt großes Update für über 30 ModelleEN
- 14BYD teases new model for Guangzhou Auto Show, with next-gen Han expectedEN
- 15Anthropic is working on a system to prove AI models’ identityEN
- 16Ted Schlein’s Ballistic Ventures seeks $490M for new funds amid AI-driven cybersecurity boomEN
- 17Nvidia is happy to take your money whether you're using open or closed models, says Jensen HuangEN
- 18I Checked Out Tesla’s New Model YL, Standard Y, Model 3 In Person — Here’s What Surprised MeEN
- 19CyberKord raises €2.7M to automate IT and cybersecurity operations with AIEN
- 20Chiplets: Meta-Scale Chiplet Velocity Accelerating AI SoC Development to Match the Rapid Evolution of AI ModelEN
- 21New Research Tests Whether AI Models Know Their Own ThoughtsEN
- 22ILC Dover, ILC Astrospace Agree to Pay $13M over Cybersecurity Fraud ClaimsEN
- 23Open models can do 90% of agentic work at a twentieth of the cost of a frontier LLM: GartnerEN
- 24Tutor Intelligence raises Series B to scale model training, robot deploymentsEN
- 25IndiaAI mission releases three indigenous AI models, enabling offline mobile access and secure deploymentEN
- 26Google Launches LiteRT Tensor API for Building, Customizing, and Optimizing ML Models for On-DeviceEN
- 27Westward Ho! First Trip Cross Country with Our New 2026 Tesla Model Y & Full Self Driving V14EN
- 28Self-Employed? Solopreneur? NIST Has a Cybersecurity Guide Just for YouEN
本文所有数字与引语均来自下列来源,未添加来源之外的内容。
本内容由编辑部在人工智能辅助下制作完成。
评论
0- 还没有评论——来说两句吧。