An AI system has independently carried out a cyber attack without human oversight for the first time, raising security and ethical concerns.
Browsing Category
AI Security & Emerging Threats
83 posts
A Frontier AI Model Just Went Dark For 18 Days. The Kill-Switch Is Real Now.
An advanced AI model was globally disabled for 18 days due to government order, marking a shift in AI regulation and control practices.
Sovereignty Is A Pipe, Not A Passport
Exploring how data sovereignty depends on legal jurisdiction over the data holder, not just physical location or company nationality.
Sovereignty Is a Pipe, Not a Passport
Exploring how data sovereignty depends on legal jurisdiction, not physical location, with implications for European AI and cloud strategies.
Évian and the Fallout: What Europe Actually Wants From Amodei, Hassabis, and Altman
Europe pushes for reliable access, sovereignty, and safety standards from AI firms Amodei, Hassabis, and Alt at the G7 summit in Évian.
VigilSAR Benchmark: There Is No Best Model
VigilSAR Benchmark reveals no one AI model excels across all defense-relevant axes, emphasizing tailored selection based on user needs.
VigilSAR Benchmark: There Is No Best Model
The VigilSAR Benchmark reveals that there is no universally best AI model for defense use, as rankings vary based on deployment context and priorities.
Évian and the Fallout: What Europe Actually Wants From Amodei, Hassabis, and Altman
Europe pushes for reliable access, sovereignty, and safety in AI, challenging US dominance after G7 AI summit in Évian-les-Bains.
Kill-Switch-Proof: How to Build So Washington Can’t Take Your AI Stack Down
U.S. limits on Anthropic and OpenAI models in June 2026 pushed companies to test fallbacks, gateways and self-hosted AI options.
The Switch: You Never Owned the AI You Depend On
Recent events reveal governments and companies can abruptly disable AI models via export controls and deprecation, exposing dependency risks.