Post by J.B. Branch

I joined a panel discussion on AI governance focused on answering the question of how can governments procure AI systems that are safe, reliable, and trustworthy? The conversation brought together an outstanding group of experts, including Rebecca Kern of The Capitol Forum, Clara Langevin of Federation of American Scientists, Jae Yeon Kim of The University of North Carolina at Chapel Hill, and Quinn Anex-Ries of Center for Democracy & Technology. My contribution focused on a recent paper I authored for the Federation of American Scientists examining AI procurement in K–12 education. The paper argues that policymakers should focus on governance before deployment. Specifically, I recommend: establishing procurement guardrails for high-risk AI systems; requiring Algorithmic Impact Assessments before deployment; restricting predictive-policing and law-enforcement-derived technologies in schools (schools should be a safe place for ALL kids); creating ongoing monitoring and transparency requirements (including accessible information for parents to make informed decisions/challenges to AI-enabled outcomes); providing technical assistance and vendor accountability mechanisms; and investing in leadership-level training so school officials can effectively oversee these technologies. The TLDR: Innovation and accountability must advance together. Most importantly, schools should not be forced to choose between technological progress and student safety. Read the paper on the FAS website here: https://lnkd.in/g28mp9-9 #EdTech #AIGovernance #EducationPolicy #PublicPolicy #TechPolicy