Remodelling internet infrastructure: A first look at platform governance in the era of ChatGPT

Abstract

OpenAI’s ChatGPT, the hybrid private company–nonprofit’s latest artificial intelligence (AI) project,arrives just as the Government of Canada attempts to pass its own response to AI: the ArtificialIntelligence and Data Act (AIDA). Amidst ongoing debates over ChatGPT and its growing connectionsto major platforms, we draw attention to the complex web of policy concerns this very new technologyraises and urge greater consideration of the information commons as a key policy frame to understandAI chatbots and the large-language models (LLMs) used to train them. ChatGPT could not exist withoutthe collective production of resources to support and maintain these commons, and its exploitation ofthem will only continue as OpenAI and its competitors search for ways to monetize chatbots.