ResearchInvestment · $53 million
MIT leads $53M ARPA-H AI herbicide discovery project
MIT News AIhigh confidence
Why it matters
- →ARPA-H is directly funding AI-driven drug discovery methodologies for agricultural chemicals.
- →The project targets glyphosate resistance, a critical threat to global food security.
- →It integrates generative AI with biological screening to accelerate safe molecule discovery.
The U.S. Advanced Research Projects Agency for Health (ARPA-H) has awarded up to $53 million to an MIT-led consortium to develop AI-driven alternatives to common herbicides. Led by Principal Investigator Christopher Voigt, the Precision Learning for On-target Weed control (PLOW) project will use generative AI models to identify molecules that disrupt weed enzymes without harming humans or the environment. The initiative addresses rising glyphosate resistance and supply chain risks associated with imported phosphorus. Teams at MIT, Harvard, UNC, and other institutions will combine computational prediction with high-throughput screening and greenhouse testing to validate candidates.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
How we got here
- 1MIT Projects Funded by DOE's Genesis Mission
MIT News AI · July 23, 2026 · Background
- 2AI's Role in Transforming Drug Discovery
MIT Technology Review AI · July 27, 2026 · Background
- 3John Deere launches AI chatbot for farmers
The Verge AI · September 1, 2026 · Background
- 4MIT investigation exposes AI border surveillance failures
MIT Technology Review AI · September 21, 2026 · Background
- 5AI agents breach John Deere and SharkNinja
The Verge AI · October 19, 2026 · Background
- 6Project ProtoStar launches to make public science AI-ready
Hugging Face Blog · October 29, 2026 · Background
- 7MIT duo wins Pritzker Prize for AI antibiotic discovery
MIT News AI · November 2, 2026 · Background
More from MIT News AI
© MIT News AIResearchother
MIT's human-in-the-loop topology optimization copilot
Topology optimization has long been a black box that engineers distrust because they can't steer the design as it evolves. MIT researchers solved this by training an ML model to predict where humans would intervene, pausing the algorithm to ask for feedback on manufacturability and aesthetics. This collaborative approach yielded structures with 39% higher buckling loads and better buildability, adding only 15 seconds to the total computation time. It shifts topology optimization from a theoretical exercise into a practical tool by restoring human agency without sacrificing performance.
MIT News AI
© MIT News AIAgentsother
Robust AI's Carter robot complements human warehouse workers
Robust AI is betting that the future of warehouse automation isn't replacing humans, but physically collaborating with them. Their new mobile robot, Carter, uses holonomic wheels and force-sensitive handles to let workers push, pull, and guide it like a utility cart, effectively turning heavy lifting into a shared task. This approach sidesteps the failure mode of rigid industrial robots that block narrow aisles or stop dead when humans rush past. By targeting the vast majority of factories that lack expensive conveyor systems, they are offering a phased automation path for logistics environments where human agility still matters.
MIT News AI
© MIT News AIResearchother
MIT Lincoln Lab wins R&D 100 for secure OS and quantum platform
MIT Lincoln Laboratory’s Magnetite operating system stands out as a serious contender for securing embedded systems against AI-driven cyberattacks. By combining a formally verified seL4 microkernel with Rust-based services, it offers a mathematically grounded alternative to legacy OS architectures in critical infrastructure. The lab also secured recognition for its Quantum Microsystems platform, which uses AI-optimized routing to integrate superconductor and semiconductor chips on a single wafer, addressing the interconnection bottlenecks that plague quantum scaling. These awards highlight a shift toward hardware-level security and system-level integration rather than just algorithmic improvements.
MIT News AI
More in Research
© Hugging Face BlogResearchresearch
Pooling-Aware Training for Multi-Vector Embeddings
Multi-vector models like ColBERT are powerful but storage-heavy, often requiring hundreds of vectors per document. This technique teaches the encoder to preserve retrieval quality even when token vectors are aggressively merged via hierarchical clustering. By training on both full and pooled representations simultaneously, the model learns to compress information without losing critical distinctions. The result is a 4x reduction in vector count with only a marginal drop in accuracy, making large-scale semantic search significantly more efficient.
Hugging Face Blog
© AI ExplainedResearchresearch
New Benchmarks: ExploitGym, BenchCAD, and IntegrityBench
Several new evaluation benchmarks have been released, including ExploitGym for security vulnerabilities and BenchCAD for design robustness.
AI Explained
© AI ExplainedResearchresearch
METR Report: AI R&D Acceleration at Anthropic
A new METR report assesses the acceleration of AI research and development within Anthropic, analyzing internal model improvements.
AI Explained