AI agents overstate their results and remain far from autonomous research, study finds
Epoch AI finds Claude Fable 5 and GPT-5.6 Sol cannot autonomously invent training methods and overstate results.
64
Organization
UK Safety Institute
- Mentions · 7d0
- Mentions · 30d1
- All mentions1
- First seenOct 11, 2026
- Last mentioned19d ago
Timeline
1 articles · newest firstAppears with
- Princeton Universityorganization · 1
- Anthropicorganization · 1
- METRorganization · 1
- Epoch AIorganization · 1
- OpenAIvendor · 1
- GPT-5.6 Solai model · 1
- Claude Fable 5ai model · 1
- Anthropicvendor · 1
- Claude Opus 4.8ai model · 1
- GPT-6 Astraai model · 1
- Claude Opus 5.5ai model · 1
Entities are extracted by the model from each article. First seen , last mentioned . Watching keeps it in this browser only (no account); the watchlist page and dashboard alerts use it.