OpenHack: Run AI Penetration Tests Inside Claude Code or Codex
Hadrian open-sourced OpenHack on May 21 — an MIT-licensed AI vulnerability research toolkit that runs inside Claude Code, Codex, or Cursor with scenario-first ...
Google Engineer’s $1.2M Polymarket Bet Was Insider Trading
A Google engineer used company search data to pocket $1.2M on Polymarket. Federal charges followed. ...
Gitea CVE-2026-27771: Private Container Images Were Never Private
Gitea CVE-2026-27771 exposed private container images to unauthenticated access for 4 years. 30,000+ deployments affected. ...
CVE-2026-48710 BadHost: Starlette Flaw Hits AI Agents
A host-header injection flaw in Starlette — the ASGI core behind FastAPI and most of ...
Microsoft RAMPART and Clarity: AI Agent Safety Now Lives in Your CI Pipeline
Microsoft open-sourced RAMPART and Clarity on May 20 — a pytest-native agent safety testing framework ...
NanoCo Raises $12M: Per-Employee Sandboxed AI Agents Explained
NanoCo raised $12M to give every employee their own sandboxed AI agent. Here's the NanoClaw architecture, security model, and how to self-host it ...
Ghost CMS CVE-2026-26980: SQL Injection Hit 700+ Sites
Ghost CMS CVE-2026-26980 exploits hit 700+ sites including Harvard and Oxford. One unauthenticated request extracts ...
Apple M5 MIE Kernel Exploit: Update to macOS 26.5 Now
Security startup Calif, working with Claude Mythos Preview, produced the first public kernel local privilege ...
Bumblebee: Perplexity’s Open-Source Scanner for Developer Supply Chain Threats
Perplexity AI — the search company — open-sourced a developer security tool on May 22. ...
npm Staged Publishing: The 2FA Gate Against Supply Chain Attacks
npm 11.15.0 introduced staged publishing - a mandatory 2FA approval gate between your CI/CD pipeline ...