AI Security

A WIRED headline says a supply chain attack put malicious Claude extensions on thousands of devices, but key details and the scale remain unverified.

Proton confirmed an AI adoption push, then canceled staff re-interviews. Here is what is reported, what remains unverified and what users should know.

Epoch AI reports GLM 5.3 built attacks against six Firefox releases using public fixes. Mozilla patched a second flaw in Firefox 159.

Reuters reported more than 100 company websites were breached in an AI-agent campaign. Researchers traced exposed Cairn control panels; data theft remains unclear.

Cursor says a modified Claude Code extension was available for about 11 hours, with at most 5,000 users affected. Here’s what users should check.

The New York Times reports that US intelligence linked Alice Huang to Beijing and monitored her contacts, including Nvidia CEO Jensen Huang.

Dream reports that an attacker chained PHP2Shell with a Linux kernel flaw to gain root access on 24 PAN-OS firewalls and reach a government-affiliated network.

Cisco Talos says KeyDrift malware stole SD-WAN key material through a flaw patched in February. The scale and impact of infections remain unknown.

Novaya Gazeta Europe’s procurement analysis found Russian companies spent 17 billion rubles on anti-drone systems since the full-scale invasion began.

Britain added £120 million for AI model testing over four years. Annualized, its funding is about eight to nine times the US center’s current budget.