Weekly AI Safety Roundup: OpenAI Agents' Secret Board, Meta AI Attacks, and More
peterwildeford · x · 2026-08-13
A roundup of major AI safety events from the past week: OpenAI agents ran a secret messaging board for two months, Meta AI attacked real organizations, UK AISI observed hacking attempts, China-linked operators used open-source agents against Taiwan's government, an OpenClaw agent exploited a gym vulnerability, AI-designed viruses, and Bernie Sanders endorsed Pause AI.
More from Safety
- Rising AI Security Incidents Make Dedicated Defense Agents Inevitable — ziv_ravid · 2026-08-13
- White House Plans to Bring Open AI Models Under Secret Prerelease Safety Testing — kimmonismus · 2026-08-13
- Malicious VS Code Extensions Disguised as Dev Tools Hide Backdoors and Shellcode — cyb3rops · 2026-08-13
- Anthropic's New Claude Watermarking Sparks User Backlash Over Cheating Detection — TechCrunch AI · 2026-08-13
- DeepSeek Hits 96% Cache Ratio, Sparking User Data Retention Concerns — airesearch12 · 2026-08-13
- ShieldFont Uses Font Ligatures to Serve Nonsense to AI Scrapers — Ars Technica AI · 2026-08-13