PRIME PANEL at the Scottish AI Summit April 2023

Professor Gina Netto | 22nd November 2023

In April 2023, PRIME PI Professor Gina Netto (Heriot Watt University (HWU)) chaired a panel which combined inputs from research team members Dr. Aunam Quyoum (University of Glasgow), Professor Lynne Baillie (HWU) and community organisation partners Colin Lee (Chief Executive, CEMVO Scotland) and Charles Kwaku-Odoi (Chief Officer, Caribbean & African Health Network (CAHN)) at the Scottish AI Summit. The Summit, held in Glasgow at the University of Strathclyde, brought together key notes and inputs from experts in Scotland’s AI community to facilitate conversations with diverse stakeholders.

Following an introduction to the main goals of the PRIME project, its multi-disciplinary, cross-institutional, and cross-sectoral nature, panel members considered a wide range of areas, including the need for greater awareness of the barriers faced by individuals from minoritised ethnic communities in engaging with online services and the importance of effective engagement with individuals from these communities in the design, development and deployment of digitalized services. The need for greater understanding of ethnicity data as well as clarity around the purpose of collecting such data, was also discussed, along with how processes of data collection and analysis could be improved.  Panel members also emphasized the potential and actual bias and exclusion in systems which were designed primarily around the needs of white males. Consequently, it is important to enable users among minoritised ethnic communities to hold service providers and others to account. The need for better understanding of how minoritised ethnic communities could be effectively engaged in service design and review was also stressed. Both service providers and web developers were encouraged to collaborate with the research team to co-design and co-produce better online services and privacy enhancing technologies to improve the safety of minoritised ethnic communities online and avoid the perpetuation of bias within existing services.

A short polling activity with members of the audience indicated that while a high proportion of attendees were collecting ethnicity data, a much smaller proportion were testing services with individuals from minoritised ethnic communities, and an even smaller proportion were designing their services in consultation with these individuals, suggesting that greater awareness of the importance of testing and designing services in consultation with these users is needed. Among the areas which were further discussed were the value of an intersectional approach in engaging with service users, the scope for AI to empower individuals from these communities and the need for greater understanding of  the promotion of ethnic diversity of the workforce at various levels in the field of AI. A video of the session can be found HERE