The food hygiene watchdog has implemented a variety of both ‘generative and traditional’ automation technologies, including trials of a mobile app supporting the work of frontline safety inspectors examining businesses

The Food Standards Agency is using a range of artificial intelligence technologies to support its work, including frontline inspections.

According to comments made by Andrew Gwynne – who formerly served as a minister in the Department of Health and Social Care but has now been removed and suspended from the Labour party over offensive WhatsApp messages – the “FSA is using AI, both traditional and generative, for a range of purposes”.

This includes trials of generative AI technologies for use by inspectors working to examine companies providing or preparing food.

“For generative AI, we are piloting its use in our front-line services in the field by using mobile-based AI applications,” Gwynne said. “The goal of this is to streamline our inspection of meat businesses by having AI help collate notes during the inspection process, which will allow uniformity in reporting and improve data quality. We aim for this to improve the existing method, which involves inspectors carrying large amounts of equipment while taking written, paper-based, observations.”

Additionally, generative systems are being used in the internal operations of the food regulator, the recently deposed minister said.

Related content

• Food allergy reporting tool aims to improve policymaking
• Food Standards Agency to connect 220 UK abattoirs​
• Defra invests in social media analysis tool to monitor public attitudes to food issues

“We have also deployed generative AI tools to improve data quality,” he added. “As most data from national and international food alert systems is unstructured text, considerable human effort has been required to extract the relevant information and then categorise it to a standardised format. The aim is to reduce the manual work required in improving data quality, which will allow colleagues to spend more time deriving insights from data rather than cleaning data, while also improving the speediness of the response.”

As well as the implementation of technologies that can independently create content from scratch, the FSA has used more “traditional” automated systems to help analyse documents.

“Using traditional AI, we have focused on pattern detection for food risk identification using, and developing, approaches to extract and structure information contained in documents, from shipping manifests to webpages,” Gwynne said. “We aim to see food safety and authenticity risks before the food lands on the United Kingdom’s shores.”

The FSA is a watchdog which works across England, Wales and Northern Ireland to oversee the safety and hygiene of food. It operates as a non-ministerial department.

Gwynne’s comments were made in response to a written parliamentary question from Conservative MP John Hayes