Best AI Alignment Research Newsletters in 2026
NewsletterFIT tracks 3 newsletters that cover AI alignment research as a primary topic, all with posts in the last 90 days. By publicly shown subscriber count, the largest are David Friedman’s Substack (5K) and LAB10 (2K).
Updated Nov 13, 2026 · sorted by publicly shown subscribers · methodology
- 1David Friedman’s Substack5K subscribers
SubstackSeveral times a week · last post 11 days ago
From recent posts: David Friedman's Substack publishes weekly essays on game economics, Bayesian statistics, and legal theory using World of Warcraft as a case study, alongside occasional reader response posts and legal reform arguments.
Likely audience: Economics students
- 2LAB102K subscribers
SubstackSeveral times a week · last post 10 days ago
From recent posts: LAB10 publishes technical analysis of AI agents, Anthropic's business and IPO filings, and formal software verification using LLMs. It is aimed at engineers, analysts, and technical readers tracking AI safety, token…
Likely audience: data analysts
- 3
SubstackWeekly · last post 10 days ago
From recent posts: Entropy Bonus publishes essays on AI topics including existential risk, alignment research, neural network interpretability, large language model training, and critiques of the stochastic parrot metaphor. It is written…
Likely audience: AI safety researchers
Latest posts from these newsletters
- DeepSeek ahora…LAB10
- AI made…Entropy Bonus
- AI, Doom…David Friedman’s Substack
- Tu equipo…LAB10
- Tu KV…LAB10
- The Romantic…Entropy Bonus
- Anthropic sale…LAB10
- Another Kind…David Friedman’s Substack