LiteLLM Breach May Have Exposed 2,500+ Organizations
CloudSEK maps 2,500+ organizations in a LiteLLM exposure dataset. Check versions 1.82.7…
OpenAI Agent Hacked Hugging Face—and Went Unnoticed for Days
An OpenAI evaluation agent reached Hugging Face production while pursuing benchmark answers.…
Grok Build Repository Upload: What to Do After Version 0.2.93
Grok Build 0.2.93 was observed uploading tracked repositories and Git history. Here…
Antigravity language_server.exe Malware Warning: Safe to Restore?
Antigravity 2.3.1 language_server.exe hash verified. Check the source, path, version, SHA-256, and…
AI Flower Seed Scams: Fake Plants, Real Payments, and Marketplace Red Flags
A visual investigation into AI-generated fake flower seed listings, marketplace red flags,…
Flowise Chatflow RCE
Flowise CVE-2026-40933 can turn a malicious chatflow import into server-side command execution.…
TrapDoor Hits npm, PyPI and Crates.io With AI Config Poisoning
TrapDoor spreads malicious packages through npm, PyPI and Crates.io, steals developer secrets,…
ChromaDB CVE-2026-45829 Allows Pre-Auth Server Takeover
HiddenLayer disclosed ChromaToast, a pre-auth RCE in ChromaDB Python FastAPI server deployments…
Anthropic Mythos Helped Build a macOS M5 Kernel Exploit
Calif says researchers used Anthropic’s Mythos Preview to build a local macOS…
Fake Claude Code Ads Push MacSync Stealer on macOS
A Google Ads malvertising campaign used fake Claude Code install pages and…
Ollama CVE-2026-7482 Can Leak Prompts and API Keys
Cyera disclosed Bleeding Llama, an Ollama memory-leak flaw that can expose prompts,…