Anthropic Red Team: Small AI Models May Soon Self-Replicate on IoT Devices
Anthropic's Frontier Red Team published research on October 12, 2026, reporting that a 9-billion-parameter model ran inference on eight of nine hardware types tested, and models of 30 billion parameters or fewer can already complete many individual replication steps. The team projects small models will reach the required capability level by the end of 2026.
- 9B model ran inference on 8 of 9 hardware types tested
- 15 $55 Android TV boxes served 0.39 tokens/s on a 26B model
- Only GLM-5.2 achieved full replication: 8 of 8 nodes in 4 of 5 trials
- Team projects small models reach the needed level by end of 2026
Read next
AI