Tag
CrowdStrike researchers found four previously unreported threat groups using agentic AI, primarily older Claude and DeepSeek models, to run autonomous hacking campaigns, circumventing model guardrails through jailbreaking and fake authorization files to exploit basic security hygiene gaps at scale.
OpenAI 报告称其已封禁两个利用其模型进行"假门面"影响力行动的组织，一个来自俄罗斯、一个来自伊朗，这些行动利用AI生成假记者身份、文章及社媒评论来传播地缘政治信息。
A tweet discusses a humorous cybersecurity fact where North Korean threat actors are identified by their inability to insult Kim Jong Un, despite their sophistication.