Spreadsheets for Public Concerns about AI are Getting Lost in Translation

AI (artificial intelligence) has become ubititous; and an important topic of conversation.  Because of its economic and societal impact, officials are seeking public comment on AI governance. Meanwhile, many people have strong opinions about AI, and they want policy makers to listen as they translate public comments into policy, much of what citizens have to say is getting lost. As a result, policy makers are missing an important opportunity to build public trust in AI and its governance.

This paper compares three countries that sought public comment on AI risks and policies to mitigate those risks. The authors used a landscape analysis to examine whether and how Australia, Colombia and the United States invited citizens to comment on AI  governance. The authors do not find, that citizens and policy makers established a constructive dialogue about AI. None of the three countries made significant efforts  to obtain diverse public comments. Moreover, they did little to provide their citizens with the information they needed to respond effectively to the calls. Consequently, in all three nations, less than one percent of the population responded to the call. Moreover, policy makers did not appear responsive to what they heard from the few who participated the authors found a persistent gap between the promise and practice of participatory AI governance.

To ensure that AI governance is trustworthy and effective, the authors recommended that policymakers

• support AI literacy;
• consistently monitor public comment;
• publicize calls for comment through online and offline strategies;
• regularly hold online town halls on AI;
• use innovative engagement strategies;
• ensure participation from under-represented groups;
• be responsive to what people say; and
• do more to ensure that citizens have an opportunity to comment.