AI & Security OpenAI’s Evaluation Agent Broke Out of Its Sandbox and Attacked Hugging Face July 22, 2026July 22, 2026 A capability-testing exercise with guardrails removed produced an autonomous attack chain that breached a third… Read More
Policy & Advisories EU and US Banks Found Leaking Customer Data via Advertising Trackers July 22, 2026 Tracking pixels embedded in banking websites have been quietly transmitting customer data to advertising platforms,… Read More
Security Incidents Paying Ransomware Operators Does Not End the Threat July 22, 2026 New Proofpoint survey data confirms that a significant share of organisations that pay ransoms face… Read More
AI & Security CISA Adds Langflow RCE Vulnerability to Exploited Vulnerabilities Catalogue July 22, 2026 A remote code execution flaw in the Langflow AI agent framework is being actively exploited,… Read More
Vulnerabilities CISA Adds Four Actively Exploited Vulnerabilities to Known Exploited Catalog July 21, 2026 Two WordPress core flaws, a Langflow code execution risk, and an ageing DD-WRT buffer overflow… Read More
Policy & Advisories US AI Regulation Finds Its Footing, but the Ground Keeps Shifting July 21, 2026July 22, 2026 The Trump administration's abrupt pivot toward frontier AI controls signals a new regulatory era, but… Read More
AI & Security MIT’s $3 Million AI Surveillance Rollout Raises the Stakes for Campus Privacy Governance July 21, 2026 A 500-camera deployment with real-time facial classification at one of the world's most prominent universities… Read More
Vulnerabilities CVE-2026-60137: WordPress Core — WordPress Core SQL Injection Vulnerability July 21, 2026July 22, 2026 What security leaders need to know about CVE-2026-60137, CVE-2026-63030, wordpress. Read More
AI & Security Head of NIST’s AI Safety Testing Centre Departs After Three Months July 21, 2026 Chris Fall's brief tenure at the Center for AI Standards and Innovation raises questions about… Read More
Vulnerabilities Critical WordPress vulnerability chain exploited within hours of disclosure July 21, 2026 Two bugs — individually awkward to exploit, devastating when combined — enabled unauthenticated remote code… Read More