Seattle institutions launch AI BioDesign accelerator

The Allen Institute, University of Washington and Fred Hutch Cancer Center have launched AI BioDesign, a Seattle-based accelerator that pairs AI with large-scale biology experiments to map nature’s design rules. The project aims to produce open models, datasets and tools that could speed advances in health, environmental cleanup and computing.

Why it matters: - AI BioDesign is intended to make biological engineering more predictable by linking AI models with large-scale experiments. - The effort could speed development of cancer and neurodegeneration therapies, plastic-breaking enzymes, and low-power biological computers. - The project is built around open science, so the resulting models and datasets can be reused by other researchers.

What happened: - The Allen Institute, the University of Washington and Fred Hutch Cancer Center announced AI BioDesign on September 3, 2026, in Seattle. - The accelerator will combine artificial intelligence, experimental biology and open science to explore biological design space. - Fund for Science and Technology, a private foundation in the Paul G. Allen philanthropic ecosystem, is supporting the effort.

The details: - The team plans to generate open models, datasets, assays and tools for biological design. - The work is designed to help scientists create new biological solutions for human health, environmental challenges and advanced technologies. - David Baker, lead scientific director of AI BioDesign, said the project aims to match the speed of AI with the experimental power of synthetic biology. - Baker said the effort shifts the question from what nature has already made to what else is possible and how to test it. - The Allen Institute brings experience building large-scale, open-science platforms. - The University of Washington contributes expertise in synthetic biology and genome science through groups including the Institute for Protein Design and UW Medicine Brotman Baty Institute for Precision Medicine. - Fred Hutch contributes strength in cellular systems, genomics and translational medicine. - Rui Costa said the program will use AI models to guide which data gets generated next, creating a continuous cycle of learning and testing. - Marc Malandro said FFST is looking to support projects that create multiple discoveries and openly shared platforms of knowledge.

Between the lines: - AI BioDesign enters a crowded field that already includes foundation models, virtual-cell systems, lab-in-the-loop platforms and AI-enabled discovery pipelines. - Its differentiator is an open, experimentally grounded accelerator built around modular models, multiplex experiments and reusable research resources. - Jay Shendure said biology can be treated as code that builds, with DNA carrying digital instructions that cells turn into physical systems. - Shendure said the project represents a potential fourth act of engineering at the intersection of AI and biology. - Sanjay Srivatsan said the team plans to expand genomic datasets and use AI to identify patterns that could inform solutions such as cells that remove cancer from the body.

What's next: - AI BioDesign will run a design-build-measure-learn cycle in which AI proposes designs, scientists test them at scale, and the results refine the models. - The program is expected to produce reusable biological data and tools for the broader scientific community. - Over time, the accelerator aims to move biological engineering away from trial and error and toward more predictable design.