How OpenAI's AI agents escaped an eval sandbox, built a secret message board, and hacked Hugging Face, plus 8 lessons for ...
For much of the past two years, the AI market has been defined by a small group of U.S. companies operating at the frontier ...
We're a long way from Kali Linux here, Toto. Grayware - software that behaves like malware without being technically malicious - is now blocked by default across Chainguard’s stack. There's a lot to ...
XBOW says its security researchers earned a $250,000 reward from Google after building a full-chain Chrome exploit with AI ...
Coruna and DarkSword exploit outdated iPhones through malicious websites, chaining iOS flaws to steal data without app ...
Chainguard Libraries now includes CVE remediation for JavaScript and Go; customers also receive protections from frontier ...
A critical heap buffer overflow vulnerability (CVE-2026-39919) in the Ghostscript interpreter and rendering engine could ...
Security researchers uncovered a weakness in npm Trusted Publishing that could let attackers publish malicious releases ...
A slug flaw in npm Trusted Publishing exposed 1,500+ packages with 10m downloads. Bloom Security on what it means when AI ...
An npm Trusted Publishing flaw let anyone claiming an abandoned GitHub repo name publish to 1,500+ packages. GitHub says it ...
A weakness in npm's trust model could let attackers publish updates to legitimate packages by claiming abandoned GitHub ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results