Artificial intelligence is helping Bayer compress what has traditionally been one of agriculture’s longest innovation cycles, cutting the time required to develop some new crop varieties by more than half.
The company has reduced development timelines from as much as a decade to between two and five years in certain breeding programs, according to Tom Eickhoff, Vice President and Head of Research and Insights at Bayer Crop Science.
The acceleration offers a glimpse into how AI is reshaping agricultural R&D, from seed development to crop protection and regulatory processes.
“We’re using AI to model breeding crosses, predict performance across different environments and make decisions much faster,” Eickhoff said in an interview with The AgriBiz.
Plant breeding has historically relied on years of field testing. Researchers would plant, harvest and evaluate successive generations of seeds, repeating the process over multiple growing seasons before determining whether a new variety was ready for commercialization.
AI is changing that equation. By modeling genetic combinations and predicting how plants are likely to perform under different soil and weather conditions, algorithms can eliminate part of the trial-and-error process.
The result is fewer experiments, faster decisions and a higher likelihood of success.
“This is literally reducing product development timelines from eight to 10 years to two to five years,” Eickhoff said. “It’s not simply about moving faster. It’s about making better decisions and increasing the probability of getting it right.”
The technology’s impact extends well beyond breeding. Eickhoff said AI is influencing nearly every stage of Bayer’s research and development pipeline.
“From early discovery and development to how we interact with regulators and register products, and even how our field teams and employees perform their daily work,” he said.
Data as a Competitive Advantage
For Bayer, the key differentiator is not AI itself, Eickhoff said, but the vast amount of scientific data the company has accumulated over decades.
The company maintains extensive databases covering genetics, crop protection products and field trials conducted around the world.
“AI helps us make sense of that data complexity, but it also enables us to go a step further in understanding genomes and identifying potential pest targets, for example,” he said.
Gene editing has become another pillar of that strategy. Bayer is using the technology alongside AI to develop products that more precisely address farmers’ needs and bring them to market faster.
Eickhoff pointed to partnerships in the field, including Bayer’s collaboration with Pairwise, the company known for launching the first gene-edited food product in the US.
CropKey Platform
In crop protection, Bayer’s AI-driven R&D efforts are centered on a platform known as CropKey.
The system is designed to accelerate the discovery of new molecules for weed, pest and disease control while also helping researchers assess product safety early in the development process.
“From the earliest stages, we can identify products that are not only highly effective but also highly safe, with a strong probability of delivering the outcomes we want for farmers,” Eickhoff said.
The approach is already generating commercial candidates. Bayer used the platform to develop Icafolin, a new herbicide mode of action targeting weeds, an issue of growing concern for Brazilian farmers.
The company estimates the product could generate about €750 million in global sales and expects to begin launching the herbicide in Brazil from 2028.
Despite AI’s expanding role across research and development, Eickhoff said the technology cannot replace one critical element: direct engagement with farmers.
Maintaining close contact with growers remains essential both for understanding their most pressing challenges and for supporting the adoption of new technologies, he said.
“It’s very important for us to stay close to farmers because we’re bringing more technology to them, faster than at any other time in history,” Eickhoff said.
This story was translated with the assistance of artificial intelligence
