The pilot is designed to enhance and improve the agency’s ability to quickly and efficiently identify imported seafood products that may pose a threat to public health. This is especially important since the United States imports upwards of 94 percent of its seafood supply.
In 2019, the FDA launched the first phase of the pilot, an analytical proof of concept, to examine the use of machine learning (ML) to target violative seafood shipments. Machine learning is a type of AI that makes it possible to rapidly analyse data, automatically identifying connections and patterns in data that people or even the agency’s current rules-based screening system cannot see. The analysis demonstrated the potential for AI to assist FDA in ensuring the security and safety of the nation’s food supply, specifically imported seafood.
As part of the FDA’s “New Era of Smarter Food Safety” initiative, the agency is undertaking a cross-cutting effort to leverage its use of AI-based technologies. The data from this new pilot program will be studied and used to evaluate the utility of AI in support of import targeting, which may ultimately help implement an AI model to target high-risk seafood products.
The new pilot programme, which is scheduled to run from 1 February to 31 July will help the agency not only gain valuable experience with new powerful AI-enabled technology but also add to the tools used to determine compliance with regulatory requirements and speed up detection of public health threats.
Following completion of the pilot, FDA will communicate its findings to promote transparency and facilitate dialogue on how new and emerging technologies can be harnessed to solve complex public health challenges.