FAR.AI
Lab Blog · Safety & Research · far.ai
AI safety research lab FAR.AI on jailbreaks, deception and risk.
Summary written from the source’s own About page · checked 2026-11-21
What changed: FAR.AI’s newest post in its feed is from Nov 10, 2026. It posted about 0.5 a week over the last 4 weeks, after none in the 4 weeks before. Listed on AI News Bay since Nov 21, 2026.
Latest Posts
Titles link to the original on far.ai. We never copy article text.
- A few lines of text bypass Kimi K3.1 Pro's safeguards | FAR.AI (opens in a new tab)
- SynthDeception: Realistic, Multi-Domain Agentic Scenarios for Studying Falsehood Detection Probes | FAR.AI (opens in a new tab)
- AI-enabled terrorism: Addressing radicalization and CBRN risks in the age of LLMs | FAR.AI (opens in a new tab)
- Jailbreaking Qoder’s Cyber Safeguards | FAR.AI (opens in a new tab)
- Persuasion Undermining Control: Can AI Talk its Way Out of Human Control? | FAR.AI (opens in a new tab)
Publishing Activity
Posts per week over the last 12 weeks, from its feed
Guides to AI News Sources
- AI Newsletters for Developers, Grouped by What You BuildAI News Sources · 5 min read · Nov 22, 2026
- The Best AI Newsletters for BeginnersAI News Sources · 6 min read · Nov 21, 2026
Similar Sources
Same main topic: Safety
Robert Miles AI Safety
Videos about AI safety research, for everyone.
Future of Life Institute
Non-profit working to reduce extreme risks from AI and other transformative tech.
METR
Updates and write-ups on measuring frontier AI capabilities and risks.
Steady: Posted in 12 of the last 12 months; first post 2023AI Alignment Forum
Community blog for technical AI alignment research.
How these numbers are made: we read this source’s public feed every 6 hours. Posting rate, last post, topic mix and “first to report” come from that feed and from how we group stories across all sources. We never show ratings or reviews. Read the methodology →