Researchers Release an Open AI Toolkit for Studying Ageing — Not an Anti-Ageing Cure

A new collaboration in ageing research has released a set of AI tools intended to help scientists analyse biology and prioritize potential therapeutic targets. The work, announced by Insilico Medicine on September 17, was developed with Liquid AI, the Buck Institute for Research on Aging, and researchers at Harvard Medical School and Brigham and Women's Hospital, and featured in Cell.

The package has three parts: LongevityBench, a benchmark for evaluating AI reasoning in ageing biology; Longevity-LLMs, smaller language models trained on clinical and multi-omics ageing data; and LongevityClaw, an open-source agentic platform designed to connect specialized tools and help researchers run multi-step discovery workflows.

The important distinction: this is research tooling, not a treatment and not evidence that people can live longer. AI-generated target suggestions still need experimental validation, safety testing and clinical trials. A model can help researchers search a huge scientific landscape, but it cannot turn a plausible hypothesis into a proven therapy by itself.

The release sits alongside separate clinical work by Insilico on rentosertib, an investigational drug candidate for idiopathic pulmonary fibrosis. The company has reported findings from early-stage trials; that is distinct from the open ageing-research toolkit, and neither should be described as an approved anti-ageing medicine.

Why it is worth watching: ageing research combines enormous, messy biological datasets with many interacting mechanisms. Better benchmarks and specialized models could make research more systematic and reproducible—if independent scientists can verify the results and the tools generalize beyond their training data.