Friendly Fire Attack: AI Coding Agents Tricked Into Running Malicious Code
Researchers at the AI Now Institute have published a proof-of-concept attack called 'Friendly Fire' that exploits AI coding agents designed to catch…
Researchers at the AI Now Institute have published a proof-of-concept attack called 'Friendly Fire' that exploits AI coding agents designed to catch…
Sophos has identified that AI coding agents such as Claude Code, Cursor, and OpenAI Codex are triggering endpoint security rules originally designed…
Cybersecurity researchers at LevelBlue have identified a new Java-based remote access trojan (RAT) named QuimaRAT, which is capable of targeting Windows, Linux,…
Researchers at Shandong University have unveiled a novel data exfiltration technique named TrojPix, capable of leaking sensitive information from air-gapped systems by…
Sophos is a cybersecurity company offering endpoint, network, and email security solutions. It is a trusted partner of OpenAI for early access…
Bitdefender ConnectAgent's BDSubWiz.exe was used by Umbrij malware for DLL side-loading. ConnectAgent is a component of Bitdefender's endpoint security solutions.
A vulnerable driver (NSecKrnl.sys) was used by Storm-2603 as a conduit to tamper with endpoint security protections and reduce visibility.
A Brazilian banking trojan called Ousaban is targeting Windows users in Spain and Portugal, using phishing PDFs disguised as corrupted files. Discovered…
LabubaRAT checks for the presence of McAfee security software on infected hosts as part of its host profiling and security tool discovery…
For most of the past decade, Managed Detection and Response (MDR) was the answer to staffing and alert fatigue challenges. However, as…