Ken Underhill

Cybercriminals are increasingly recruiting insiders to bypass security controls across banks, telecoms, and technology firms.

Researchers demonstrated how AI safety approval prompts can be manipulated to trigger remote code execution.

A critical zero-day flaw is being actively exploited to remotely compromise more than 125,000 WatchGuard Firebox firewalls.

A phishing campaign targeting HubSpot users bypassed email defenses by abusing trusted platforms and authenticated infrastructure.

OpenAI has launched GPT-5.2-Codex, an agentic coding model that boosts real-world software engineering and AI-powered vulnerability research.

Amazon uncovered a North Korean IT infiltrator through keystroke latency, highlighting risks in remote hiring and the need for stronger identity controls.

Buyer Shifts, Growth Opportunities, and How Marketers Win in a New Security Era

Q4 2025 attacks show early AI agents are already expanding the attack surface, pushing enterprises to rethink AI security for 2026.

A misconfigured childcare CRM exposed 140,000 records linking parents and children.

Moonwalk++ bypasses EDR using Windows call stacks.

The GhostPoster campaign hid malware inside Firefox extension icons, infecting tens of thousands of users through trusted add-ons.

SoundCloud confirmed a breach that exposed user data through an internal dashboard, potentially affecting millions of accounts.