That’s the prediction of Ethan Soloviev, chief innovation officer of HowGood, a sustainability intelligence platform that works across the food system. He understands the task at hand for the food industry is to feed some 10 billion people by 2050—not only growing it but getting it to people.

“The lowest-hanging fruit for AI is how it will improve the food and agricultural supply chain with rapid increases in efficiency and productivity,” Soloviev says. “AI can do the math for us.”

Using AI to get your products from point A to point B is a growing solution to logistical hurdles, but in no other industry does it feel as nuanced as the food supply chain. That supply chain includes everything from natural agricultural and weather-related challenges to grow ingredients to inventory management and product shelf life: The end consumer needs that item to stay fresh long enough to cook it and eat it, be it at home or at a food service establishment. 

It’s a $2.3 trillion industry in the U.S. alone that involves hundreds of corporations and thousands of individuals. The end product affects, well, everyone.

“Foodstuffs are less forgiving than your average package,” says Erik Nieves, cofounder and CEO of Plus One Robotics. For one, they have an expiration date. Many are temperature controlled. Some must be handled with care. And all are highly regulated by government agencies. That’s on top of typical consumer-good challenges like inventory management, demand forecasting, and warehouse efficiency. Then there is the task of physically transporting it all to the far corners of the country. 

Because these complications exist at the root of food production, Nieves explains, he has seen AI greatly reduce the time-to-shelf for several products. Part of that is with robotics, like his, that automate packaging systems in warehouses with the help of machine learning and 3D computer vision. The robots can hang out longer in a cold freezer to package temperature-controlled goods and also handle more manual labor than a human, even with a forklift. They are getting pretty good, he says, at detecting different types of fruit and adjusting their gripper strength to avoid bruising a ripe pear.

But it goes beyond mechanical arms. Also in the food warehouse, Amazon uses AI-driven sensors to detect potential food safety hazards, such as contamination and spoilage, by monitoring the temperature and humidity of storage facilities. 

Amazon Web Services (AWS) uses similar tech in AI-powered image recognition systems and machine vision technologies to inspect food products for defects, according to Justin Honaman, head of worldwide retail and consumer goods go-to-market growth. Combined, this sets up an “AI-based traceability system that can track the movement of food products from farm to fork,” he says, and is critical for notifying suppliers, retailers, and consumers of recalls resulting from foodborne illness. 

And there are endless other applications, as more AI adaptation takes place, Honaman says.

“The food industry is an excellent source of large and complex data sources with highly complex and multidisciplinary challenges, making it a prime application for artificial intelligence,” says Bryan Hitchcock, chief science and technology officer at the Institute of Food Technologists. 

One such impactful application is consumer demand. AI is improving recommendations on how much of a specific product a grocery store needs to order, to reduce food waste created by shipping excess products that won’t be sold. By analyzing historical sales data, AI is giving food distributors more insight into what is selling when—and informing its purchase orders accordingly. A 2022 study by the World Wildlife Fund found that AI software offered a 14.8% reduction in food waste per grocery store; merchandising tech companies like Trax and Shelf Engine work heavily in this space, and major food companies are in on the action, too. Kraft Heinz partnered with Microsoft to do this with its portfolio of more than 200 brands.

Hitchcock is particularly excited about how AI can work alongside existing, and powerful, technologies in the food industry, like nuclear magnetic resonance machines that help scientists analyze chemical compositions in foods; biotech tools to modify crops; food processing technologies to preserve nutritional qualities; and the many climate-focused platforms to enable agricultural and food production companies to optimize practices and work toward sustainability goals. 

At Pangea Global Technologies, Bryan Fried is an advocate of indoor farming as such a solution, thanks to his work in lighting large-scale horticulture projects. In a controlled, indoor environment, AI plays a crucial role in optimizing the growth atmosphere and fine-tuning crop results. Fried, who serves as the LED lighting company’s chairman and CEO, says that “when a crop is planted, there is virtual certainty that it will yield the desired result,” which has significant potential in solving food insecurity in so-called food deserts and augmenting traditional agriculture amid climate change. It may also reduce the need to transport food over long distances to remote or arid locations. 

“The use of AI is just beginning to be applied,” Fried says. “All of this will increase the availability of food, especially what we consider healthy food, to any place.”

When Hitchcock thinks about the future of AI and food, he suggests there is risk, too. AI is a tool that can be used to deliver on the priority of improving food for everyone, but it can also work against that goal. HowGood’s Soloviev notes the “countless variables” that must be considered and the number of individuals working across the food system, who may not all be aligned in improving the impact of the food system on the planet.

“With a technology as adaptable as AI, our biggest risk is the failure of the imagination,” Hitchcock says. “If we can’t imagine the possibilities of how this technology can be used, for both the good and the bad, we won’t be able to prepare for what might challenge us tomorrow.”