In FDA’s mandate to regulate the safety and nutrition of roughly 80% of the US food supply, it is turning to artificial intelligence and state regulators to help with some of the heavy lifting.
The FDA is continually strengthening its Human Foods Program (HFP), which was launched in 2024 and now focuses on three priority areas: food chemical safety, nutrition and microbiological food safety, said FDA Deputy Commissioner for Food Donald Prater in his keynote address at the IFT FIRST show in Chicago.
Prater explained that the program’s vision aims to “ensure that food serves as a vehicle for wellness” and “promote the health and wellness of the American public through science-based approaches to prevent foodborne illness, reduce diet-related chronic disease and ensure the chemicals in our food are safe.”
“That’s a big job, and we’re committed to using our resources in the most efficient and effective ways that we can,” he said.
Strategic priorities at FDA
The FDA is setting ambitious goals, including work with artificial intelligence and machine learning to prevent foodborne illness, Prater explained.
The agency also is strengthening oversight of ingredients and additives, in addition to working to reduce the prevalence of diet-related chronic diseases, he added.
“We’ve already begun a rigorous review of the range of food additives and their safety, and we’ve implemented a plan to remove petroleum-based dyes from America’s food,” he said.
Addressing the growing concern about the potential dangers of ultra-processed foods (UPF) “given their association with numerous chronic diseases,” is another top priority for the FDA, Prater said.
Bridge Project: New federal-state model
Prater touted the FDA’s Better Regulatory Inspections for Dynamic Government Efficiency (BRIDGE) Project, a multi-phase program launched a year ago with the goal of modernizing domestic food facility inspections.
Calling it one of FDA’s “most ambitious projects,” the project largely aims to transition low-risk inspections to the states, Prater said.
“Under this model, we envision a new paradigm to further enhance the integrated federal-state food safety system that relies more heavily on our state partners to conduct routine domestic oversight activities, including routine inspections and sampling.”
He said the FDA will continue to conduct inspections for non-routine and high-risk inspections, as well as foreign inspections.
“This is not a reduction in oversight; it’s a reallocation of FDA resources toward higher-risk, more complex work,” Prater said. “What we’re doing in this administration, though, is making sure that chronic risks also get the attention that they deserve, alongside more acute risks like foodborne illness.”
FDA is working with a handful of states to implement the state-led inspection program in the next four years, Prater said. That means training inspectors to comply with federal standards.
“We want to ensure that an inspection in Utah is the same as an inspection in Maine,” he said.
Steven Mandernach, executive director of the Association of Food and Drug Officials (AFDO), warned regulators in April that the shift must come with federal investment, according to an article in Food Safety News.
Mandernach told the US House of Representatives Committee on Energy and Commerce Subcommittee on Health that it would be difficult to execute the program equitably across states without financial backing and staffing.
Artificial intelligence safety assist
Artificial intelligence and machine learning is becoming a central function of the FDA’s efforts to ensure safety in the food supply, Prater explained.
The FDA is partnering with universities and industry associations to share vast amounts of data and research that can be fed into AI models to better predict foodborne illnesses, Prater said.
Those partnerships include a joint FDA-Purdue University study of salmonella and cantaloupes, efforts with Mexico to ensure the safety of produce imports and a pilot project on leafy greens data sharing with the Western Growers Association, he said.
“As much as whole genome sequencing has been a game changer for the microbiological world. AI and machine learning is really a game changer for food safety, and so FDA leadership is championing a greater focus on AI and ensuring rapid enterprise development of AI tools that are being shared across FDA,” he said.
AI and machine learning models are also being used to predict seafood shipments that are in violation of US law, he said.
“Given that so much of our nation’s seafood comes from international suppliers, learning to predict which shipments into the US might be violative provides an opportunity to significantly improve the efficiency of our inspection capability,” he said.
