THREAD: AI is increasingly being used to make decisions about human welfare. In a new Day One proposal, Kevin Wei proposes three key actions that the Federal Government can take to promote the transparency and accountability of AI. https://www.dayoneproject.org/post/creating-transparency-and-fairness-in-automated-decision-systems-for-administrative-agencies

Action 1: Agencies that administer benefits programs should develop robust in-house technical capacity, as well as strong cultures centering data transparency and accountability.

Action 2: The White House should issue a Presidential Memorandum requiring that agencies submit proposed AI or automated decision systems to a notice-and-comment process before adoption.

Action 3: Congress should pass comprehensive legislation to regulate the use of AI and automated decision systems across domains including immigration and law enforcement, benefits programs, and others.

Kevin developed his proposal through a Day One Accelerator. Day One Accelerators guide a diverse cohort through a process to turn their policy idea into a catalytic proposal for action. Learn more here: https://www.dayoneproject.org/accelerators