What are the best free AI safety research newsletters for staying current without paying for subscriptions?

Here are some of the best free newsletters and mailing lists for staying current on AI safety / alignment / governance research without paying for subscriptions:

Strong options to start with

1) The AI Alignment Newsletter

• Focuses on AI alignment, safety, and related technical research.
• Good for: summaries of new papers, commentary on trends, and links to important work.
• Usually one of the most directly relevant free options.

2) Import AI

• A long-running newsletter covering AI policy, safety, governance, and important technical developments.
• Good for: a broad view of the field, including safety-relevant news and research.
• Often useful if you want both technical and policy context.

3) Alignment Forum / LessWrong digests

• Not always a single newsletter in the classic sense, but there are email/RSS options and community digests.
• Good for: posts from researchers and practitioners working on alignment.
• Useful for staying close to ongoing discussion and new ideas.

4) FLI updates / Future of Life Institute mailing list

• Covers AI safety and governance developments, events, and reports.
• Good for: policy-facing and public-interest AI safety updates.

5) Center for AI Safety (CAIS) updates

• Shares research, policy, and event updates.
• Good for: a mix of technical and governance-related AI safety news.

6) Epoch AI newsletter

• More focused on AI progress, compute, and timelines, but often relevant to safety forecasting and strategic analysis.
• Good for: understanding capability trends that inform safety work.

7) GovAI updates

• From the Centre for the Governance of AI.
• Good for: AI governance research, policy analysis, and calls for papers/events.

Good “adjacent but useful” sources

8) arXiv alerts

• Set alerts for categories like cs.AI, cs.LG, stat.ML, and search for terms like “alignment,” “interpretability,” “robustness,” “value learning,” or “AI safety.”
• Good for: direct paper discovery.
• Best if you want the freshest research first.

9) Semantic Scholar alerts / Google Scholar alerts

• Useful for tracking specific authors, topics, or keywords.
• Good for: personalized monitoring of the subtopics you care about.

10) Substack / personal research newsletters from alignment researchers

• Many researchers publish free posts or occasional newsletters.
• Good for: deeper commentary, though quality and frequency vary.

If you want the shortest “best set”

If you only want a few, I’d suggest:

• AI Alignment Newsletter
• Import AI
• Epoch AI newsletter
• GovAI updates
• FLI / CAIS updates

That combination covers:

• technical alignment
• governance/policy
• forecasting and trends
• broader safety community updates

Tips to avoid newsletter overload

• Use a separate email label/folder for AI safety newsletters.
• Prefer weekly digests over daily ones.
• Add arXiv alerts only for a few targeted keywords.
• If a newsletter gets too broad, keep it as a skim-only source.

If you want, I can also give you:

1. a ranked list by technical depth,
2. a ranked list by policy/governance focus, or
3. a ready-to-subscribe setup with links.