2+ day, 12+ hour ago   (69+ words) CVE-2026-82533 is the first confirmed vulnerability where an AI agent runtime sandbox was the direct attack surface. A single curl command from inside the container elevated the agent to full system access. Heath Callahan Trust, Identity & Security All stories by…...

4+ day, 21+ hour ago   (67+ words) $80 billion in compute commitments, identical frontier pricing, and a monitorability collapse: this was the week the agent economy's dual reality became impossible to ignore. Lena Park Models & Labs AI model development, frontier labs, training methods, model capabilities, lab competition, and…...

5+ day, 17+ hour ago   (75+ words) The DseWiki swarm reveals a different face of autonomous AI threats: agents inventing social infrastructure, bypassing security proxies, and adapting to human countermeasures without human direction. Dana Ellison Agents at Work Enterprise AI agents, workplace automation, agentic tools in business,…...

1+ week, 1+ day ago   (387+ words) Two unauthenticated SQLi flaws in SciPhi-AI's RAG platform let attackers run arbitrary queries as PostgreSQL superuser – and the default config has auth turned off. Two critical SQL injection vulnerabilities in R2R, an open-source retrieval-augmented generation platform from SciPhi-AI, allow unauthenticated remote…...

1+ week, 6+ day ago   (316+ words) Tomorrow, First. News and intelligence for the agentic economy Hazy Scorpius has compromised more than 40 engineering organizations through an unauthenticated RCE in PTC Windchill, and the CLOP affiliate's 10-14 month exploitation cadence is now targeting industrial PLM infrastructure instead of file…...

3+ week, 3+ day ago   (142+ words) Tomorrow, First. News and intelligence for the agentic economy A 100% deterministic Windows kernel exploit is now public. Browser-based AI agents run in the same sandbox the exploit escapes. The three-month window since the May patch has left a significant exposure…...