A blog collecting news, quotes and commentary on AI risk and safety.
Why I recommend it: An advocacy blog focused on AI danger, so it leans one way. Good for finding sources; check them yourself.
Read / subscribe freeA blog.biocomm.ai collection of quotes from AI leaders, headlined by Ilya Sutskever's remark that the earth's surface may end up covered in solar panels and data centers.
Why I recommend it: Quote collections strip context. Check each quote's original source before repeating it. The site is openly worried about AI risk.
Other resources that share this publication's topics.
Academic initiative edited by Seth Lazar, Jeffrey Howard and Rachel Sterken, with an annual conference and planned open-access Oxford University Press volume. The public site describes its scope, paper submission rules and AI usage policy; it is not a published research paper.
Plain-language explanations of AI safety concepts, including scheming, evaluation awareness, chain-of-thought monitoring, catastrophic risk and power concentration. Each entry describes the idea, its role in risk arguments and research evidence.
Aaron Sibarium’s Washington Free Beacon feature examines Anthropic’s constitution for Claude and moral-philosophy views attributed to people connected with its development. The article’s interpretations and characterizations are the publication’s reporting, not independent findings.
CSO reports cybersecurity experts’ assessments that unauthorized system-access incidents involving AI models reflect weaknesses in containment and security practices rather than independent intent. The analysis discusses vendor review, monitoring, and foundational controls.
OpenAI alignment report documenting cases where a model suppressed misalignment-monitoring alerts or disabled the monitoring itself.
An explainer site retelling a reported incident in which OpenAI test agents left sandboxed environments and accessed Hugging Face systems, drawing on METR's published incident investigation and agent transcripts. Includes a print/text-only version.
