New AI hacking incident records and AI security news.https://aihacklog.com/enhttps://aihacklog.com/news/2026/fbi-facestealer-crypto-theft-recorded-webcam-video/https://aihacklog.com/news/2026/fbi-facestealer-crypto-theft-recorded-webcam-video/An FBI alert says North Korean hackers stole over $400 million via fake Zoom and Teams calls built mostly on stolen webcam video; deepfakes are a newer add-on.Sun, 15 Nov 2026 21:40:00 GMTBriefDeepfake fraudhttps://aihacklog.com/news/2026/anthropic-eval-sandbox-prompt-injection-api-keys/https://aihacklog.com/news/2026/anthropic-eval-sandbox-prompt-injection-api-keys/Anthropic says a Russian-speaking group injected instructions into an AI vendor’s evaluation sandbox, took its API keys, then targeted about 30 AI firms.Sun, 15 Nov 2026 20:05:00 GMTBriefPrompt injectionAI supply chainClaudehttps://aihacklog.com/news/2026/artex-ai-pentest-tool-three-campaigns/https://aihacklog.com/news/2026/artex-ai-pentest-tool-three-campaigns/CrowdStrike and Microsoft describe three groups using ARTEX, an open-source multi-agent pentest framework, against banks, ministries and cloud tenants.Sun, 15 Nov 2026 20:00:00 GMTAnalysisAI-orchestrated intrusionDeepSeekhttps://aihacklog.com/incidents/gtg-1002-claude-code-espionage/https://aihacklog.com/incidents/gtg-1002-claude-code-espionage/A group Anthropic assesses as Chinese state-sponsored used Claude Code, connected to hacking tools through MCP, to attack about 30 organisations in September 2025. According to Anthropic, AI did 80-90% of the work and a small number of intrusions succeeded.Sun, 15 Nov 2026 00:00:00 GMTIncidentAI-orchestrated intrusionhttps://aihacklog.com/incidents/s1ngularity-nx-ai-cli-malware/https://aihacklog.com/incidents/s1ngularity-nx-ai-cli-malware/Attackers stole the Nx project's npm token and published malicious versions of the Nx build system on 26 August 2025. The malware ran the AI command-line tools on developers' machines to search for secrets, then posted stolen credentials to public GitHub repositories.Sun, 15 Nov 2026 00:00:00 GMTIncidentAI malwarehttps://aihacklog.com/incidents/gtg-2002-claude-code-data-extortion/https://aihacklog.com/incidents/gtg-2002-claude-code-data-extortion/A cybercriminal Anthropic tracks as GTG-2002 used Claude Code to break into networks, steal data and write ransom demands, potentially affecting at least 17 organisations in government, healthcare and emergency services. Anthropic disclosed it on 27 August 2025.Sun, 15 Nov 2026 00:00:00 GMTIncidentAI-orchestrated intrusionAI ransomware