other · first seen 6 Oct 2026
AI safety
Also: AI alignment, agent safety guarantees
Research on preventing learned systems from causing harmful outcomes through failures, attacks, and unsafe decisions.
Mentions per week
PapersCodeDiscussion
Earliest sources
- Toward Alignment Scaling Laws: A Framework and First Preregistered MeasurementsPapers · 6 Oct 2026 · Jeremy Canale
- AI Safety Considerations for Agents With Limited Time to ActPapers · 7 Oct 2026 · Leo Zeitler, Jack Richings, Victoria Nockles
- OpenAI cannot make AI safe on its own [pdf]Discussion · 8 Oct 2026 · Philpax · 23 points
- Towards Safe World Models: Risks, Defenses, and EvaluationPapers · 9 Oct 2026 · Xiaoma Wang, Litian Zhang, Xinze Shi, Zhengxuan Hu, Zihan Yang, Mingqi Ren, You…
- Alignment via GeneralizationPapers · 25 Oct 2026 · Shai Shalev-Shwartz, Amnon Shashua
- AI Alignment Is Neither Solvable nor SufficientPapers · 9 Nov 2026 · Gerrit Großmann
Latest sources
- AI Alignment Is Neither Solvable nor SufficientPapers · 9 Nov 2026 · Gerrit Großmann
- Alignment via GeneralizationPapers · 25 Oct 2026 · Shai Shalev-Shwartz, Amnon Shashua
- Towards Safe World Models: Risks, Defenses, and EvaluationPapers · 9 Oct 2026 · Xiaoma Wang, Litian Zhang, Xinze Shi, Zhengxuan Hu, Zihan Yang, Mingqi Ren, You…
- OpenAI cannot make AI safe on its own [pdf]Discussion · 8 Oct 2026 · Philpax · 23 points
- AI Safety Considerations for Agents With Limited Time to ActPapers · 7 Oct 2026 · Leo Zeitler, Jack Richings, Victoria Nockles
- Toward Alignment Scaling Laws: A Framework and First Preregistered MeasurementsPapers · 6 Oct 2026 · Jeremy Canale