Assured SI for Autonomous Biology

Call for proposals opens: October 2026

The "Trust, Assuredness, and Validation of Super Intelligence (SI)-Driven Autonomous Biological Laboratories" campaign, led by Maia Kapur and Timothy Boe, addresses a growing gap in autonomous science: self-driving laboratories can now propose and run experiments faster than scientists can verify them, and the machine-learning models steering them are often most confident where they are wrong.

Through this Computing, Analytics, and Modeling Community Science Campaign, EMSL will partner with invited researchers who run autonomous or LLM-driven experimental loops in the life sciences to independently assess the trustworthiness of their workflows.

Participants will contribute experimental records, including model predictions, uncertainties, selected designs, and measured outcomes, which EMSL will evaluate using lightweight trust diagnostics applied directly to those records, while also measuring the run-to-run reliability of LLM-based experiment-planning tools. Validated on EMSL's own autonomous platform (Anaerobic Microbial Phenotyping Platform) and on participants’ systems, the campaign will deliver an openly available diagnostic suite, a community reporting standard for auditable autonomous experiments, a shared cross-lab reference for interpreting results, and a co-authored best-practices publication.

The call for proposals opens in October 2026, with work expected to begin in December 2026. 

Stay tuned for more information, including an opportunity to submit a proposal.