MIT gets $53M to build AI that weeds out harmful chemicals

An MIT-led team has secured up to $53 million from the U.S. Advanced Research Projects Agency for Health to tackle a growing agricultural crisis. The project, called PLOW, aims to use generative AI to discover new, safer weed killers as traditional herbicides face resistance mutations and raise health concerns.

The research will focus on developing AI models that can predict molecules—either chemical compounds or RNA-based treatments—that target the same plant enzymes as current herbicides but with fewer health risks. This could reduce U.S. dependency on imported phosphorus used in glyphosate and help farmers maintain crop yields without exposing workers and consumers to potential carcinogens. The initiative represents a practical, AI-driven approach to solving a real-world problem that balances agricultural efficiency with public health.