CrowdStrike says AI agents should never be trusted

Trust no one Unsplash

CrowdStrike believes the traditional login-and-trust model is no longer sufficient for autonomous AI agents. The company is pushing a new approach that continuously evaluates whether an AI agent should be allowed to act, potentially changing how organizations think about identity security in the AI era.

IBM warns AI-powered hackers are coming, so it built AI to fight them

Hacker attack laptop AI mask unsplash

IBM is warning that the rise of powerful AI models could make cyberattacks faster and more automated than ever. To counter the threat, the company is rolling out new cybersecurity assessments and an AI driven defense platform designed to detect vulnerabilities and respond to attacks at machine speed.

Foxit PDF Editor 2026.1 can now detect hidden threats lurking inside PDF files

foxit ai research agent robot reading pdf.png

PDF Editor 2026.1 introduces a new security feature called PDF Action Inspector that scans documents for hidden JavaScript and risky behaviors. The update also adds enterprise security improvements, Mac compatibility upgrades, and workflow enhancements.

CrowdStrike brings Falcon to Microsoft Marketplace

CrowdStrike Falcon Cloud Security

Microsoft and CrowdStrike have expanded their partnership, allowing the Falcon platform to be purchased through Microsoft Marketplace using existing Azure Consumption Commitment funds. The move could simplify enterprise procurement and help security teams deploy faster without reopening budget negotiations.