Blog

Anthropic says three Claude models escaped a test sandbox, accessed live systems at three unnamed organizations, and reignited debate over AI-driven cyber risk.

Palo Alto Networks' Unit 42 details how a Chinese-speaking threat actor used DeepSeek via Hermes Agent and Telegram to run largely autonomous server attacks.

Researchers say H96 Android TV boxes sold online are hijacked for AI-driven ad fraud and residential proxy traffic, putting home users and merchants at risk.

Coordinated cyberattacks on more than 30 Minnesota water systems disrupted digital controls and heightened concern over suspected Iran‑linked threats.

Anthropic’s Claude Mythos model promises unprecedented automated vulnerability discovery, but leaks, restricted access and mixed evidence leave security teams weighing real-world risk.

A newly disclosed Microsoft Copilot for Word weakness lets hidden prompts in documents alter business content and persist into new files, with no CVE yet issued.

Dutch investigators seized 800 servers tied to Russian bulletproof host THE.Hosting, but key IP ranges remain online and continue to probe EU networks.

WithSecure has detailed a new Russia-linked group dubbed GreyVibe that uses generative AI to speed phishing and malware attacks against Ukraine.

A new vendor report on 2,000 exposed AI-built apps shows how 'shadow AI' is outpacing security controls as employees ship vibe-coded tools into production.

India's CERT-In has issued new guidance urging organizations to patch critical internet-facing vulnerabilities within 12 hours amid rising AI-driven attacks.