The World Health Organization (WHO) is asking consumers to help counter the “unethical” online promotion of unhealthy products by sending in pictures of products that will “train” its artificial intelligence (AI) tool.

The new online tool, developed by the WHO European Office for the Prevention and Control of noncommunicable diseases (NCDs), enables people to upload images of harmful products, effectively training the AI tool to better identify violations of established marketing guidelines.

The harmful products include tobacco, alcohol, foods high in salt, trans fats, and sugars as well as instances of marketing of breastmilk substitutes (BMS) and foods for infants and young children.

“We have created our AI tool to support governments to develop new ways to monitor and protect the health of children,” said Dr Kremlin Wickramasinghe, WHO/Europe’s regional adviser on nutrition, physical activity and obesity.

“We also want to encourage a wide movement among the public to take part in this work if they want to improve our children’s future, to promote healthier choices for a better life.”

The WHO said: “Whenever you come across online or offline advertisements promoting any products that could negatively impact children or undermine efforts to protect, promote and support breastfeeding, take a screenshot or picture and upload it to the WHO tool using the link provided on this page.”

The WHO is asking the public to upload images related to the marketing of any milks that are specifically marketed for feeding infants and young children up to the age of three years, including infant and follow-up formulas and drinks for young children; any other product marketed or otherwise represented as suitable for feeding infants up to the age of six months, as anything fed to a baby before the age of six months, including water, will replac breastmilk; feeding bottles and teats; and commercially produced foods that are marketed as being suitable for infants and young children from the age of six months to 36 months.

ASA continues to monitor flouting of formula marketing rules in the UK

The UK watchdog, the Advertising Standards Authority (ASA) recently ruled that online adverts promoting infant formula products breached standards aimed at protecting breastfeeding.

The four Google ads from UK retailer Boots, which appeared on April 11, 2023, showcased infant formula products from Aptamil, Hipp Organic, Kendamil (First Infant Ready to Feed), and Cow & Gate.

The ads, accompanied by images of the product packaging, raised concerns because they were marketing communications for infant formula, which is strictly prohibited.

Boots UK Ltd, which said they were fully aware of the legal responsibilities relating to infant formula, apologised for the error that had led to the ads appearing.

They had taken immediate steps to manually remove all their infant formula advertising from Google as well as from other search engines.

Vietnam’s AI-powered Virtual Violations Detector (VIVID)

The WHO is not the only one taking action to root out firms flouting infant formula marketing regulations.

Vietnam’s FHI Solutions Innovation Incubator recently unveiled another AI-powered tool, designed to scan the internet, in particular social media, to identify violations of the International Code of Marketing of Breast-milk Substitutes.

The Virtual Violations Detector (VIVID) builds on these legal measures to auto-detect Code violations through AI learning, making the scanning of huge numbers of pages and posts possible.

“During the development phase of VIVID Code, we consulted partners and Code advocates at national level, studied existing Code monitoring data and identified companies and sites that are potentially more active in digital marketing,” said Constance Ching, technical consultant for Code Advocacy and Implementation in the Alive & Thrive East Asia Pacific regional office.

“As a result, a certain number of companies, based on how active they appear on social media and the internet, were selected for auto-detection monitoring.”

Since its launch in October 2022, VIVID has detected 23,853 violations – despite only operating in selec countries (Australia, Hong Kong, Malaysia, New Zealand, Nigeria, Philippines, Singapore, UK and USA) and monitoring selec companies.