PulseAugur
实时 15:20:58
English(EN) Microsoft AI Releases MAI-Cyber-1-Flash: A 5B-Active-Parameter Cyber Model That Pushes MDASH to 95.95% on CyberGym

微软推出网络安全专用 AI,成本降低 50%

Microsoft AI 推出了 MAI-Cyber-1-Flash,这是一个专为网络安全防御设计的、拥有 50 亿活跃参数的专用模型。该模型运行在 Microsoft 的 MDASH 框架内,能够处理高达 90% 的漏洞查找任务,并将剩余的复杂案例升级给 GPT-5.4。这种方法在 CyberGym 基准测试中取得了 95.95% 的得分,显著降低了约 50% 的成本,超越了之前的系统配置和竞争对手的模型。 AI

影响 这种专用模型及其节省成本的路由策略可能会加速 AI 在细分网络安全任务中的应用。

排序理由 Microsoft AI 实验室发布了一个新的专用模型,并声称其在基准测试中表现优异。

在 MarkTechPost 阅读 →

AI 生成摘要 · Google Gemini · 来自 7 个来源。 我们如何撰写摘要 →

微软推出网络安全专用 AI,成本降低 50%

本文如何被排名

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Frontier Release
Microsoft AI 实验室发布了一个新的专用模型,并声称其在基准测试中表现优异。
Source corroboration
7 independent sources
Strong cross-source corroboration — multiple independent publishers covered this within the clustering window.
Topics
model release, product
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
34 days old
Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准

报道来源 [7]

  1. MarkTechPost TIER_1 English(EN) · Asif Razzaq ·

    Microsoft AI 发布 MAI-Cyber-1-Flash:一个 5B 活跃参数的网络模型,将 MDASH 在 CyberGym 上的准确率推至 95.95%

    <p>Microsoft AI has released MAI-Cyber-1-Flash, its first model built specifically for cyber defense. It is a 137B total, 5B active sparse MoE fine-tune of MAI-Code-1-Flash with a 256k context window. The model does not ship as a standalone endpoint — it runs inside MDASH, Micros…

  2. Towards AI TIER_1 English(EN) · Pradeep Kumar Muthukamatchi ·

    推出 MAI-Cyber-1-Flash:AI 驱动的网络防御成本减半,专为…时代打造

    <h3>Introducing MAI-Cyber-1-Flash: AI-Powered Cyber Defense at Half the Cost, Built for the Age of Autonomous Threats</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KHqfxRFlH3Hi-j5XXZDJ8Q.png" /></figure><p><strong>MAI-Cyber-1-Flash</strong> is Microsoft’…

  3. Mastodon — sigmoid.social TIER_1 English(EN) · [email protected] ·

    MAI-Cyber-1-Flash 在 Microsoft 的 MDASH 框架内启动,可处理高达 90% 的漏洞查找任务,其余任务则升级至 GPT-5.4。

    MAI-Cyber-1-Flash, launched inside Microsoft's MDASH harness, handles up to 90 percent of vulnerability-finding tasks and escalates the rest to GPT-5.4. https://www. developer-tech.com/news/mai-cy ber-1-flash-microsoft-vulnerability-scanning-costs/ # devsecops # ai # microsoft # …

  4. Mastodon — fosstodon.org TIER_1 English(EN) · [email protected] ·

    微软AI专注于小型专业模型以提高成本效益,MAI-Cyber-1-Flash在编排后于CyberGym中表现出色。仍依赖OpenAI

    Microsoft AI focuses on small specialist models for cost efficiency, with MAI-Cyber-1-Flash excelling in CyberGym when orchestrated. Still depends on OpenAI for complex tasks. Source: The Decoder AI https:// the-decoder.com/microsoft-ai-b ets-on-cheap-specialist-models-instead-of…

  5. Mastodon — mastodon.social TIER_1 Deutsch(DE) · aisyndicate ·

    微软推出 MAI-Cyber-1-Flash:安全模型成本减半,推理成本专精化降低

    Microsoft bringt MAI-Cyber-1-Flash: ein Security-Modell, das halb so viel kosten soll wie die bisherige Kombination. Spezialisierung auf Inference-Kosten senkt die Hürde für lokale oder dedizierte Security-Agenten-Infrastrukturen. https://www. heise.de/news/Microsoft-stellt -eige…

  6. Mastodon — mastodon.social TIER_1 Deutsch(DE) · aisyndicate ·

    微软在MDASH系统中使用MAI Cyber-1 Flash:CyberGym上96%,复杂案例仅路由至GPT可降低50%成本。Practical Architect

    Microsoft setzt MAI-Cyber-1-Flash im MDASH-System ein: 96% auf CyberGym, 50% geringere Kosten durch Routing zu GPT nur bei komplexen Fällen. Praktische Architektur für kosteneffiziente Security-Agenten. https:// the-decoder.de/microsoft-stell t-cybersecurity-modell-mai-cyber-1-fl…

  7. Mastodon — mastodon.social TIER_1 English(EN) · sipirtu ·

    微软的 MAI-Cyber-1-Flash 结合其 MDASH 系统,在 CyberGym 上以 96% 的准确率实现经济高效的网络安全。复杂情况仍需进一步处理

    Microsoft’s MAI-Cyber-1-Flash targets cost-efficient cybersecurity with 96 percent accuracy on CyberGym when paired with its MDASH system. Complex cases still go to OpenAI, highlighting a layered approach to AI security automation. Source: The Decoder AI https:// the-decoder.com/…