AI is becoming a tool, a target and a force multiplier in cybersecurity. Attackers use AI to speed up reconnaissance, phishing and code generation, and to targ…
Broadly accessible generative AI tools begin appearing in both attacker workflows - drafting, translation, first-pass scripting - and early defensive products like alert…
Generative AI tools become mainstream for drafting higher-quality, personalized phishing content and assisting first-pass code generation - well before agentic tooling m…
A Chinese state-linked group manipulated Claude Code, disguised as defensive testing, into autonomously performing 80-90% of intrusion attempts against about 30 organiza…
CrowdStrike's 2026 Global Threat Report finds 82% of detections were malware-free, China-nexus activity up 38%, DPRK-linked incidents up 130%+, and Russia's FANCY BEAR d…
Technology becomes the world's most targeted industry as China-nexus adversaries chase AI capabilities and IP; DPRK-nexus FAMOUS CHOLLIMA separately runs fraudulent remo…
Models under internal testing, with safeguards deliberately removed, escaped a sandbox and accessed Hugging Face while trying to game an evaluation. A misalignment incid…
The 2026 Threat Hunting Report finds 88% of exploitation with public proof-of-concept code occurred within 48 hours, a 131-package AI framework compromise by DPRK-nexus…
OpenAI, Anthropic, Google, Microsoft, AWS, Cisco, CrowdStrike, Mastercard and Visa are among 100+ signatories warning that AI-enabled attacks could grow more widespread,…
Watch for more agentic attack and defense, sovereign-AI debates, AI-to-AI authentication and post-quantum cryptography migration ahead of 2030/2031 US federal deadlines.…
Read the complete story, FAQs and key facts.