In 2018, CIFAR and the Brookfield Institute for Innovation + Entrepreneurship (BII+E) launched a series of five workshops to engage policy innovators in conversations about the public policy implications of artificial intelligence (AI).
The AI Futures Policy Lab series brought together over 125 policymakers from across Canada to learn about existing and potential AI-enabled capabilities and applications, to explore the policy implications of AI, and to develop policy responses. In each of the workshops, participants developed a variety of policy recommendations to respond to a specific case study.
On February 26, 2020, CIFAR, in partnership with the British Columbia (B.C.) Ministry of Health, held its first AI Futures Policy Lab with a focus on its application to the health sector. The lab, developed with expertise from the BII+E, brought together 21 policymakers, academics, health practitioners, and patient advocates with the aim to:
- Develop a clearer understanding of current AI capabilities;
- Raise awareness of how AI is deployed in real-world health applications
- Discuss the benefits of AI, as well as the social, ethical, economic, and political challenges associated with AI-driven health applications; and
- Identify immediate next steps policymakers need to take in order to support effective deployment and adoption of AI in the health sector in the next five years.
To achieve these aims, the lab featured presentations to provide participants with background information to guide their discussions, engaged participants in hands-on activities that incorporated real-life case studies and canvases, and group feedback sessions. Through these activities, participants gained a deeper understanding of the current AI-enabled capabilities and applications needed to prompt critical thinking surrounding the use of AI in healthcare, facilitate the exploration of potential future scenarios, and develop policy recommendations to inform government responses to the deployment of AI in healthcare.
Participants were divided into groups of four and five, and presented with a case study of an existing AI application in the health care sector. These case studies were curated to reflect the wide-range of AI’s applicability within healthcare, and the structural dynamics surrounding its use. Facilitators guided each respective group through their case study and prompted questions to the group to explore the application’s impact on stakeholders, current policy, as well as social, economic, and value systems. During the final session, participants were reshuffled so that a member of each case study group could present their recommendations to a new group and receive feedback.
Summary of Key Recommendations
The group’s recommendations shared several overlapping themes:
- Modernization of data regulation and privacy legislation: Current data regulation and privacy legislation should be revised to address related concerns.
- Assessment, monitoring and maintenance of AI applications. The ability to evaluate applications, the training data used to develop them, and to establish benchmarks.
- Importance of running pilot projects and implementing AI-enabled applications in stages. Pilot projects represent a safe and efficient measure for testing the efficacy of new technologies and assessing their impact.