Freedom of Expression in the Age of Generative AI: Between Speech Autonomyand Synthetic Manipulation

Abstract

The emergence of generative artificial intelligence (AI) systems such as ChatGPT, DALL-E, and Gemini has revolutionized how speech, creativity, and expression are produced and consumed. Yet, this transformation challenges long-standing constitutional doctrines surrounding freedom of expression and informational autonomy. The traditional notion of speech—rooted in human intent, accountability, and authorship—is being redefined as algorithmic models autonomously generate, replicate, and manipulate communicative content. This paper argues that the rise of generative AI blurs the legal boundary between speech autonomy and synthetic manipulation. By analyzing jurisprudential developments and doctrinal theories of expressive freedom, it demonstrates how algorithmic immutability creates new semi-permanent categories of expression, where bias, discrimination, and misinformation are structurally embedded. The study compares U.S. First Amendment protections with the European Union’s regulatory emphasis on dignity and democratic integrity. The article concludes that safeguarding free expression in the AI era requires doctrinal reforms: introducing algorithmic transparency, redefining editorial liability, and ensuring epistemic due process in digital speech ecosystems.