What Lila Sciences does
Lila Sciences (lila.ai) is a privately held AI-for-science company building what it calls an “operating system for science” that autonomously executes the scientific method. Lila’s core product concept is its AI Science Factory (AISF™): tightly integrated loops that combine a scientific reasoning model, tools/automation, and experimental “verifiers” so hypotheses can be generated, experiments designed and run in physical labs, and results fed back into learning in near real time. The company positions its approach as general-purpose across multiple scientific domains—explicitly including life sciences/biotech, chemistry, and materials/physical sciences—rather than narrow point solutions.
Lila’s publicly described outputs emphasize scale and breadth of experimentation. For example, it has described testing more than 950,000 mRNA sequences through its RNA discovery efforts, and claims that its platform can outperform established benchmarks in CAR-T-related work when evaluated in non-human primate settings. It also describes multi-domain “self-play with real-world experimentation,” where model proposals are validated by autonomous lab runs, and the system iterates on what it learns from those outcomes.
Commercially, Lila describes deploying its scientific superintelligence platform to “partners” and “customers,” with stated work spanning medicine/therapeutics, energy/environment, and advanced materials. Strategically, it has also signaled increasing integration with U.S. national labs and federal programs (for example, its participation in the U.S. Department of Energy’s Genesis Mission via Phase I lighthouse challenges), reflecting a roadmap to apply its closed-loop approach in higher-instrumentation, higher-scale environments.