Plate Nº 16 · recorded November 10, 2026
Biology & EvolutionReported finding
MIT-led team gets up to $53 million to seek AI-designed weed killers
An MIT-led team has secured up to $53 million to use generative AI, rapid screening and greenhouse tests to develop herbicide alternatives as common weeds evolve resistance.
By Elena Vasquez5 min read1,001 words
In brief
- MIT's PLOW team has up to $53 million in ARPA-H STREAM funding.
- More than 1 billion pounds of herbicide are used in the United States each year.
- The AI models are expected to generate billions of chemical compound or siRNA candidates.
- Initial RNAi greenhouse tests are expected to finish by the end of program year one.
- The ARPA-H funding agreement number is 140D042690030.
By Elena Vasquez
An MIT-led research team has been awarded up to $53 million to develop herbicide alternatives with generative artificial intelligence, with first greenhouse tests of initial RNA-interference (RNAi) candidates expected by the end of the program’s first year.
The award comes from the U.S. Advanced Research Projects Agency for Health (ARPA-H) through its Systems for Tracking and Resilience in Efficient Agricultural-input Management (STREAM) program.
Christopher Voigt, head of MIT’s Department of Biological Engineering, is principal investigator for the project, known as Precision Learning for On-target Weed control (PLOW).
Farmers use more than 1 billion pounds of herbicide in the United States each year to prevent significant crop loss and support regenerative farming practices that protect soil, but weeds are becoming resistant to the most common products.
Exposure to herbicides may contribute to a higher risk of cancer, metabolic and endocrine disorders, and other adverse outcomes. Metabolic disorders affect how the body processes nutrients; endocrine disorders affect hormones. Farm workers can have frequent exposure through their work, while many U.S. residents encounter small amounts through food and drinking water.
The team plans to use generative AI to discover potentially more effective herbicide alternatives developed with human health in mind.
Why does the team need new weed killers?
Some weeds have acquired genetic mutations that make common herbicides less effective and threaten crop yields.
Glyphosate works by interfering with EPSP synthase, an enzyme plants need to make several amino acids, the building blocks of proteins.
“With herbicides like glyphosate and atrazine, there’s a real crisis because of resistance mutations that are coming,” Voigt said.
He also said that combining resistance with national-security concerns about supply chains for herbicide precursors, or the ingredients used to make them, could leave the United States without agrochemicals needed for high crop yields and economical prices.
“There’s a real need for new herbicides because of resistance mutations that have been arising,” Voigt said. “There are also concerns about herbicides’ possible impacts on the environment and on humans, so it’s important to identify alternative ways of making herbicides, which as a general class of molecule are really critical to agriculture.”
The project could also reduce U.S. dependence on imported phosphorus, which is a critical ingredient in glyphosate. Its wider team is expected to include researchers from the Broad Institute of MIT and Harvard, the University of North Carolina at Chapel Hill, the University of Tennessee at Knoxville, the University of Illinois at Urbana-Champaign and Robigo.
Resistance and supply-chain pressure are pushing the search for new weed-control options.
How will the three research teams work?
The researchers are grouped into three teams aligned with STREAM Technical Areas: generative AI, high-throughput screening and greenhouse testing.
- Generate candidates. Connor Coley, the Warren K. Lewis Associate Professor of Chemical Engineering at MIT; Regina Barzilay, the School of Engineering Distinguished Professor for AI and Health at MIT; and Alexander Tropsha, the K.H. Lee Distinguished Professor in the UNC Eshelman School of Pharmacy will lead the generative-AI group.
- Define the target. The models will predict chemical compounds or small interfering RNA (siRNA), which can switch off selected genes. Candidates would disrupt the same plant enzymes targeted by glyphosate and other herbicides.
- Optimize the design. The models will try to prevent resistance from arising and favor low synthesis costs, a low risk of inducing resistance, and low risks of human harm or ecological damage.
- Generate at scale. The models are expected to produce billions of candidate molecules for a second team to test.
- Lead rapid testing. Angela Koehler, the Charles W. and Jennifer C. Johnson Professor of Biological Engineering at MIT, and Jaime Cheah, the associate director of scientific outreach and collaborative screening in the Center for the Development of Therapeutics at the Broad Institute will lead high-throughput screening.
- Feed data back. The screening team will identify molecules that hit their targets, then return all screening data to the computational approach so it can improve.
- Move to greenhouses. After candidate pools narrow, a third team will evaluate each molecule under different agricultural conditions.
- Lead greenhouse work. Neal Stewart and Scott Lenaghan, co-directors of the Center for Agricultural Synthetic Biology at the University of Tennessee at Knoxville, will lead greenhouse work.
- Complete the leadership team. Patrick Tranel, the Tom and Peggy Ainsworth Professor of Crop Sciences at the University of Illinois; Tim Willson, the Harold Kohn Distinguished Professor in the UNC Eshelman School of Pharmacy; and Andee Wallace, PhD ’20, CEO and founder of Robigo, will also lead it.
- Set the timeline. First greenhouse tests of initial RNAi candidates are expected to be complete by the end of the program’s first year.
The plan moves from billions of computer-generated candidates to rapid screening and then greenhouse trials.
Can microbes reduce agrochemical exposure?
Researchers will also develop living microbial systems that degrade agrochemicals, or crop-protection chemicals, before they reach food, soil or waterways.
Engineered bacteria, fungi and RNA molecules called srRNA will convert chemicals and added product ingredients into biomass, the organic material produced by living organisms. The aim is a scalable way to reduce environmental and human exposure.
The microbial approach could keep some agrochemicals out of food, soil and water, though that benefit still requires testing.
What remains unproven?
The announcement describes a research program, not a finished or field-proven herbicide. It reports no candidate that has completed model generation, rapid screening and greenhouse evaluation, and no field-trial result.
The health and ecological advantages remain design goals. The first greenhouse tests of initial RNAi candidates are not expected to finish until the end of year one.
The work is funded by ARPA-H under agreement number 140D042690030. The researchers state that the announcement’s views and conclusions are their own and do not represent official U.S. government policy.
The project could ease resistance and exposure problems, but its case rests on results not yet in hand.
via MIT News (Source)
More from Elena Vasquez
Show full bio
Correspondent covering money, funding and market shifts around general science news at SciBeat.
384 articles
Nearby plates
- AI Searches Millions of Enzymes to Find Pollution Fighters
- Two MIT Scientists Win 2026 Pritzker Prize for AI-Discovered Antibiotics
- NASA Will Fly Low Over Texas Farmland to Sniff Livestock Emissions
- Locally Trained AI on Gut Microbe Data Could Shield African Livestock From Heat
- MIT's CheMeleon Predicts Molecular Behavior from Scant Data