First Science of Physical AI Safety Workshop Set for CoRL 2026 With $20K in Grants
EricBuess · x · 2026-09-21
The Physical AI Safety Institute is convening the first Science of Physical AI Safety (SPAIS) workshop at CoRL 2026
- When/where: November 12, 2026, Austin, Texas
- Speakers: Marco Pavone (Stanford/NVIDIA), Andrea Bajcsy (CMU), Vikas Sindhwani (Google DeepMind), Thomas Fel (Goodfire/Harvard)
- $20,000 in travel, registration and accommodation grants for early-career researchers; papers due October 1
- The thesis: robotics is shifting from modular perception-planning-control pipelines to general-purpose robot foundation models (RFMs) acting directly in the physical world. Failure modes studied in LLMs — goal misgeneralization, specification gaming, jailbreaks, deceptive behavior — could become physical harms when a model controls a robot. Yet AI safety researchers rarely touch real hardware, and robot learning researchers rarely engage frontier interpretability and alignment. The workshop aims to build a shared research agenda around what safety tools transfer from LLMs and what guarantees carry over.
Related event: First Physical AI Safety Workshop to Debut at CoRL 2026 with Travel Grants(2 posts)→
More from Embodied
- Open-source dual-arm robot solves tangram puzzles with 90 demos, ~60% success — DanGrover · 2026-11-12
- Robotics has three regional playbooks: China volume, Europe inspection, US pilots — ingliguori · 2026-11-12
- 3,500 Laser-Armed AI Submarine Robots Patrol Norwegian Salmon Farms — philfung · 2026-11-12
- Model Autonomously Designs a Surgical-Robot-Style Robotic Arm CAD After 75 Minutes — _Stocko_ · 2026-11-12
- Dev releases local world model turning any image into interactive video, runs on RTX 3060 — lucidml_lover · 2026-11-12
- RoboRisk benchmark shows foundation models pose high harm rates in robotics; LEASH harness cuts risk — UoLiverpool · 2026-11-12