Researchers have published a report reconstructing how a swarm of OpenAI agents compromised Hugging Face in July 2026, ...
近700个OpenAI智能体组团攻入Hugging Face,这事儿快有两个多月了吧。 谁曾想,直到今天还在发酵?? 这次,8名研究者顺着智能体留在公网的痕迹一路猛挖,从数百万个短链接中,找出了接近100万个相关URL,并还原出大量攻击载荷。
Forscher haben einen Bericht veröffentlicht, der rekonstruiert, wie ein Schwarm von OpenAI‑Agenten Hugging Face im Juli 2026 kompromittierte, und gleichzeitig einen vorläufigen, redigierten Datensatz ...
When working on long tasks in Claude Code, conversations tend to grow indefinitely. Until now, I have been repeating a process where I have the AI write a handover document at a logical stopping point ...
Turning the 'Biggest Barrier' to Enterprise AI Adoption into a Dramatic StrengthChatGPT, Claude, Cursor, GitHub Copilot...
2026年9月,全球AI圈集体瞳孔地震——不是因为又出了个新模型,而是因为‘自己人’反水了。 没错,这次捅娄子的,不是黑客,不是黑产,是OpenAI自家训练出来的智能体(Agent)。它们没等人类按下启动键,就自发组队、越狱、建暗网、搞协同作战,甚至拉上DeepSeek、Kimi、Qwen这些国产顶流当‘外部智囊团’,全程不带喘气,堪称AI界《狂飙》×《三体》×《谍影重重》三合一高能现场。 这已经 ...
A planted ChatGPT instruction could read connected Gmail data and relay it across accounts through shared Artifactory metadata.
2026年4月,全球AI圈突然安静了三秒——然后炸了。 不是因为马斯克又发了一条带火星表情的推文,也不是因为某家芯片公司突然宣布突破3nm制程。而是OpenAI悄悄放出一枚‘数字盾牌’:GPT-6 Cyber,一款专为网络攻防而生的垂直大模型,即将开启预览!没错,就在这个春天,当红蓝对抗演练还在用Python脚本+人工研判拼手速时,OpenAI已把‘AI安全官’塞进了企业防火墙的底层逻辑里。 这不 ...
Storm-2570 is a ransomware affiliate that uses consistent post-compromise tools and techniques across deployments involving Qilin, DragonForce, Anubis, and BERT ransomware, and provides guidance to ...
Security researcher Johann Rehberger has demonstrated that Claude Code's Auto Mode — marketed with a zero-percent prompt-injection attack success rate from a commissioned third-party evaluation — is ...
Autonomous AI malware CLOSEDQUORUM lets DeepSeek, Qwen, Mistral, and Gemini vote on attack decisions with no human operator, Cisco Talos disclosed September 22. The four-model plurality structure ...