Project Glasswing Exposed a Vulnerability Remediation Crisis
Anthropic’s Project Glasswing proved AI can find security vulnerabilities at unprecedented scale. Claude Mythos Preview discovered thousands of zero-days across every major operating ...
Linux Kernel Disclosure Policy: No Distro Heads-Up
The Linux kernel security team’s disclosure policy ignited Hacker News today, pulling 519 points and ...
IBM Bob: Multi-Model AI Coding with Security Flaws
IBM launched Bob on April 28-30, 2026—an AI coding platform that targets enterprises, not individual ...
Rivian’s Internet Kill Switch: First EV Privacy Control
Rivian offers the first internet connectivity kill switch in a major EV, letting drivers disable ...
PyTorch Lightning Malware: 8.3M Downloads Hit by Shai-Hulud Attack
On April 30, 2026, versions 2.6.2 and 2.6.3 of PyTorch Lightning—one of the most popular ...
GitHub RCE CVE-2026-3854: 88% of Servers Still Vulnerable
Wiz disclosed critical GitHub RCE vulnerability CVE-2026-3854 on April 28. GitHub.com patched in 6 hours, but 88% of Enterprise Servers remain exposed 55 ...
OAuth Bypass: Vercel Breach Exposes Platform Secret Risk
Vercel's April 2026 breach shows how OAuth tokens bypass MFA and platform environment variables create ...
Project Glasswing: AI Finds Bugs But <1% Get Fixed
Anthropic's Project Glasswing revealed Claude Mythos Preview found thousands of zero-days, but <1% were patched. ...
Mercor AI Breach: 40,000 Voice Samples Stolen—Lawsuits
Mercor's March 2026 breach exposed 40,000 voice recordings via LiteLLM supply chain attack. Biometric data ...
SBOMs Fail as Supply Chain Attacks Surge 73%
In March 2026, Axios—one of npm’s most-used HTTP clients with 100 million weekly downloads and ...