加拿大一家公司启动了世界上第一个自主的AI网络安全系统,它学会了没有人类规则的攻击模式,能够进行实时和适应性防御。 A Canadian firm launched the world’s first autonomous AI cybersecurity system that learns attack patterns without human rules, enabling real-time, adaptive defense.

SKADI网络防御公司(SKADI Cyber Defense Corporation)是一家加拿大公司,已启动了世界上首个自主的网络安全AI, SKADI Cyber Defense Corporation, a Canadian firm, has launched the world’s first autonomous cybersecurity AI that forms its own understanding of attacks without human-defined rules. 该系统利用未经监督的学习,观察实时事件,确定在持续前侦察的模式,并建立自己的威胁概念模型。 Using unsupervised learning, the system observes real-time events, identifies patterns like reconnaissance preceding persistence, and builds its own conceptual model of threats. 它与依赖MITRE ATT&CK等框架的传统系统不同, 它即时发现新的攻击, 随时间而适应, 分享各种部署的洞察力, 创建一个集体防御网络。 Unlike traditional systems reliant on frameworks such as MITRE ATT&CK, it detects novel attacks instantly, adapts over time, and shares insights across deployments, creating a collective defense network. 该技术由语言模型的进步所促成,减少了对大型分析小组的依赖,降低了成本,标志着在可扩展的适应性网络防御方面迈出了一大步。 The technology, enabled by advances in language models, reduces reliance on large analyst teams, lowers costs, and marks a major leap in scalable, adaptive cyber defense.