The Artificial Intelligence in Public Health Toolkit: A novel resource for stakeholder engagement

Abstract

Background Artificial intelligence (AI) has considerable potential to enhance public health. People using AI systems for public health decisions, or who are affected by such decisions, may need to understand how these systems work, or articulate how much they want decision-makers to trust the system. This public engagement project, part of the Human Behaviour-Change Project, aimed to a) explore people’s views regarding trust in, and use of, AI for public health decisions and, based on that, b) create a toolkit of resources to facilitate people critically questioning the use of an AI system. Methods Six online, public engagement workshops were conducted in England in 2021 to inform the content and design of the toolkit. Twenty-four people including members of the public, public health professionals, and researchers worked with a graphic designer to create the toolkit. Results The resulting ‘AI in Public Health Toolkit’ contains resources to enable people to evaluate AI systems and provides a roadmap for the decision process, a set of suggested questions to ask about an AI system, a guide to features of good answers and a ‘personal views tool’ prompting reflection on the answers received. Participants suggested that public health decision-makers should use the Toolkit to consult people representative of those affected by the decision to recommend whether an AI system should be used in that instance. Conclusions The ‘AI in Public Health Toolkit’ has the potential to facilitate public engagement in the use of AI in public health. The Toolkit gives those developing AI-driven systems a sense of the public’s queries regarding such systems. The resources in the Toolkit can also facilitate conversations about broader AI applications to healthcare and public services.