HLS Beyond and BKC present: AI Governance and Human Alignment

In this second session of the TechReg in AI series w/ Professor Alan Raul (see February 19th and April 9th) we address the issue of how Frontier AI companies assure human control and safety. AI is positioned to be a tremendously transformative technology but it is being developed largely outside of the government’s direct control. The Administration’s current AI framework indeed assumes that  major tech companies will be largely responsible for directing and controlling the progress and governance of frontier AI. During this session we will survey how these corporate entities have set up their governance structures, instituted compliance measures (legal conformity and safety assessments, risk management frameworks), built in technical measures (evaluations, red-teaming, monitoring), and established organizational measures (risk committees, responsible scaling policies, incident response) to address such concerns. We will also explore what the current state-of-play is in efforts to identify and prevent deception and self-protection by AI models as well as enhancements of the “interpretability” of the models. Please come and join in the discussion! Registration Required. Afternoon snacks will be provided.