On moving fast and breaking things.. again: social media’s lessons for generative AI governance

Generative AI systems are increasingly being employed globally, bringing with them both tremendous promise and substantial potential for harm. As is often the case, governance initiatives are lagging behind technological diffusion. This article looks to the very recent past–the rise of social media platforms and their associated algorithmic systems–as a guide for ongoing and future conversations about the governance of generative AI. Specifically, this article identifies a set of missteps that have characterized social media governance that have direct applicability to the generative AI context. In addressing this topic, this paper will explore issues such as the valorization of innovation as a guiding governance principle, the limitations of self-regulation, and the need for a dedicated regulatory body. All of these key components of the failure of social media governance have clear applications to, and lessons for, the governance of generative AI–a context in which the stakes may be even higher.

Duke Scholars

Author Philip Michael Napoli Sanford School of Public Policy