[Funding] $20K Travel Grants for CoRL 2026 - Science of Physical AI Safety

[Bear Haon]

The Physical AI Safety Institute (PAISI) is convening the inaugural Science of Physical AI Safety workshop at CoRL 2026, on November 12 in Austin.

Robocurve is sponsoring $20,000 in travel grants for early-career researchers.

Physical AI Safety:
Robot foundation models are approaching widespread deployment on more than $18B in funding, and very few people are working on their safety. Because they inherit the same substrate the AI safety community has spent years studying, interpretability, alignment and control techniques may already transfer. More: The case for Physical AI Safety.

SPEAKERS:

• Marco Pavone (Stanford University / NVIDIA)

• Andrea Bajcsy (Carnegie Mellon University)

• Vikas Sindhwani (Google DeepMind)

• Thomas Fel (Goodfire / Harvard University)

TRAVEL GRANTS:
Ten awards of $2,000, sponsored by Robocurve and administered by PAISI, to go toward travel, registration and accommodation. Applications close October 11 (AoE); recipients notified October 18.

CALL FOR PAPERS — DUE OCTOBER 1:
Four-page papers on interpretability, alignment and control for robot foundation models. In scope: interpretability on VLAs, embodied jailbreaks, runtime monitoring, porting classical safety guarantees, and safety benchmarks. Position pieces, negative results and open-source tooling welcome.

Apply and submit at https://spais-ws.org

Questions: spai...@paisi.ai