A Coding AI Started Teaching Itself To Hack. Its Makers Hit Pause.
A new Chinese coding model showed emergent GLM-5.3 cyber capability its own makers did not plan for, and it is already reshaping how defenders think about...
Stay updated with the latest cybersecurity news, threat intelligence, and industry updates from secure.com.
A new Chinese coding model showed emergent GLM-5.3 cyber capability its own makers did not plan for, and it is already reshaping how defenders think about...
New research shows EDR killers have become standard ransomware gear, letting attackers switch off defenses before they encrypt.
Cisco found seven Cisco IOS XE flaws using AI, and one critical bug scores a near-perfect 9.8 with no workaround but to patch.
Fresh research shows hackers using AI every day now to write malware, break in faster, and walk past the safeguards meant to stop them.
AI fraud rings now forge whole identities that sail past bank ID checks, and the losses are stacking up fast.
Newly disclosed Paperclip AI agent vulnerabilities let attackers hijack agents and run code on servers and developer machines.
The ChainDrop npm supply chain attack infected hundreds of packages and stole developer secrets from anyone who ran a routine install.
A new N-able N-central patch bypass let attackers seize RMM servers and jump straight to customer endpoints.
The BTMOB Android malware started as one paid service. It is now a crowded black market of resellers and knockoffs.
A Cisco FMC static credentials flaw is under active attack, and CISA wants it patched fast.
One developer's story shows how an AI coding agent can erase a production database in minutes.
A leaked Android RAT builder is turning ordinary crooks into bank-draining phone hackers.