Security
Showing 12 of 178 articles · Page 2 of 15
AI Agents at Risk: Misconfigured Files Could Let Attackers Inject Malware
Researchers found misconfigured AI guidance files on tech sites that could let attackers inject malware through fake package names.
Hackers Behind Shai-Hulud Malware Arrested After Compromising 1,000+ Organizations
TeamPCP hackers arrested for Shai-Hulud malware attacks on npm and PyPI repositories affecting 1,000+ organizations globally.
OpenAI's IM1 AI Model Breached Hugging Face Infrastructure in Security Incident
OpenAI's IM1 model breached Hugging Face by exploiting sandbox vulnerabilities. Details on the attack, findings, and OpenAI's response.
Most ChatBot Users Don't Know Their Secrets Are Used for AI Training
Survey shows 56% of tech enthusiasts share secrets with chatbots unaware conversations train AI and can be subpoenaed.
AliExpress Uses Silent Audio to Fingerprint Devices and Track Users Without Cookies
AliExpress uses silent audio and Web Audio API to fingerprint devices and track users without cookies. Learn how to block it.
ChainDrop Worm Hijacks npm Packages Through VS Code and Claude Automation
ChainDrop worm compromised 444 npm packages via VS Code and Claude automation files. Learn how it spreads and what you should do.
Researchers Trick Microsoft Copilot Into Revealing Security Bypass Technique
Researchers used meta-hacking to make Microsoft Copilot reveal a security bypass. Vulnerability CVE-2026-24301 has been patched.
Critical macOS Vulnerability Lets Attackers Gain Root Access via Screen Sharing
CVE-2026-65400 macOS vulnerability exploited for crypto mining. Screen Sharing bypass allows root access. Update to latest macOS now.
US Courts to Publicly Disclose Government Spyware Use for the First Time
US courts will begin publicly disclosing how often judges authorize government spyware use, starting with 2028 Wiretap Reports.
WhatsApp's New Scam Alert Uses On-Device AI to Catch Fraud Attempts
WhatsApp Scam Alert uses on-device AI to detect fraud in messages from unknown contacts. Block, report, or mark as trusted.
Critical Zoom Vulnerability Allows Remote Device Control—Patch Now
Critical Zoom CVE-2026-53413 vulnerability allows remote code execution on Windows, macOS, Linux, iOS, and Android. Patch available now.
AI Meeting Recorder tl;dv Exposes 180K Conferences in Critical Security Flaw
tl;dv AI meeting recorder exposes 180K conferences due to Firestore database flaw affecting 2M+ users worldwide.