Grant: Project Grant

Background: Unhealthy food/brand marketing has a negative impact on what children eat and on their health. Canada will soon ban marketing unhealthy foods/brands to children <13 years old on television, the internet and other media. However, very little is known about the amount and types of unhealthy food/brand ads that children see online and the government does not have a way to monitor whether companies are obeying the new law. The problem is that public health has fallen behind technologically. Marketers are using artificial intelligence (AI), but public health is still relying on humans to try to find out what marketers are doing. Goals: We will develop and test the first AI system that can assess the amounts and types of food/brand marketing that is online. We will use the system to assess differences in the amount and type of food/brand marketing that is online from before to after Canada's new marketing ban comes into effect. We will also study whether companies are obeying the new law. Methods: We will use AI to identify food/brand marketing on Canadian websites, gaming applications and YouTube videos. AI will also be used to: 1) rate the healthfulness of the foods/brands that are being marketed; 2) assess the types of marketing strategies that are being used; and 3) determine whether marketing targets children. AI and marketing experts will test our system to make sure it is accurate. We will conduct statistical testing to understand how the amount and type of marketing changes from 12 months before to 12 months after Canada's marketing ban is implemented, and whether companies are obeying the new law. Significance: This AI system will provide the first scientific evidence about the amount and type of unhealthy food/brand marketing that Canadian children may see online. It will also provide a way to study whether companies are obeying the Canadian government's marketing ban. This information can help to understand how to improve child nutrition and health.