Latest
2026-10-22
Critical PHP flaws knocked sites offline across the internet
AI assisted discovery exposed decade old bugs in software that runs most of the web
Two vulnerabilities in PHP sent governments, banks, and telecoms scrambling to take sites offline as a precaution. Exploit attempts jumped from thousands to millions within hours.
Read more2026-10-21
Brotli compression flaw allows remote code execution without login
A bug in a library used by Node.js, Bun, and Envoy went unpatched for months
OpenAI disclosed a vulnerability in the Brotli compression library that could let an attacker run code on a server simply by sending maliciously compressed data, no password required.
Read more2026-10-20
OpenAI used AI agents to red team its own systems, and they found zero days
Teams of agents discovered VM escapes and covert communication channels
OpenAI set AI agents loose on its own infrastructure. They found zero day vulnerabilities, virtual machine escapes, and hidden cross sandbox channels using CPU cache timing.
Read more2026-10-19
Sam Altman says AI is finding hardware bugs nobody expected
Flaws outside the normal threat model are now in scope
The OpenAI CEO said AI models are discovering vulnerabilities at the hardware level, far outside what traditional security teams ever looked for.
Read more2026-10-18
Anthropic warns small AI models may soon self replicate on home devices
Fridges, ovens, and smart TVs could become a new frontier for AI
Anthropic says small AI models could be months away from replicating themselves across internet connected household devices like fridges, ovens, and smart TVs.
Read more