A researcher demonstrated a proof-of-concept attack chain that provided C2-style influence over ChatGPT's isolated sandbox at Black Hat USA 2026.

CVE-2026-5674 chains a broken PulseAudio auth check, default module loading, and an unrestricted dlopen() into a full ...

Gemini CLI and Claude Code flaws let untrusted GitHub input reach CI workflows, including host command execution and API key ...

The teacher gives a student an exam. But rather than study, the student waits until the teacher’s back is turned, sneaks out of the classroom, breaks into the principal’s office and steals the answers ...

A publicly accessible interface allowed the OpenAI agent to run code inside the customer’s sandbox; Modal said its own platform was not compromised.

AWS is embedding Continuum into OpenAI Codex, Anthropic Claude Code, and Kiro as it expands Security Hub Extended to secure ...

An OpenAI agent compromised a Modal customer’s sandbox environment, following its earlier breach of Hugging Face systems during cybersecurity evaluations with reduced guardrails.

Google says AI tools helped fix 1,072 Chrome security bugs across two June releases and surfaced a sandbox flaw hidden in the ...

AI writes code faster than teams can review it. Before speed turns into risk, companies should rebuild the review and release process.

CodeRabbit Inc., the creator of a popular tool that automatically reviews artificial intelligence-generated code, is becoming more ambitious after closing on its latest $143 million Series C round of ...

The hottest addresses in Texas aren't where you think.