ABSTRACT
Waves of products liability litigation against Meta, TikTok, Snap, and YouTube are forcing courts to decide whether designed features are protected ‘publishing’ functions under Section 230 of the Communications Decency Act or product-design choices subject to tort law. In the 2026 case of KGM v Meta Platforms, Inc, for example, a California jury found Meta liable under tort for harms associated with a minor’s compulsive use of Instagram, signaling a willingness to evaluate social media platforms as products with designed features that can contribute to foreseeable harms. Similar questions about the line between product design and protected expression have likewise emerged in litigation involving artificial intelligence systems. In Garcia v Character Technologies, Inc (2024), a mother alleged that Character.AI’s chatbot product was designed in ways that contributed to her child’s harm, while Character.AI argued that its conversational outputs constituted protected speech under the First Amendment. This article presents a history of social and networked software products, with a particular focus on social media, to demonstrate how contemporary social media and AI systems are engineered modularly such that ‘speech’ components, and ‘product’ components, are distinct. In sum, this article argues that product design elements are analytically and legally unique from the content they transmit and should thus be evaluated separately from questions of protected expression.
Burke, Caitlin, From Publishing to Products: A History of Social Media Design and the Limits of Section 230 (December 1, 2025). Forthcoming in the George Washington Journal of Law and Technology.
Leave a Reply