ECPR

Install the app

Install this application on your home screen for quick and easy access when you’re on the go.

Just tap Share then “Add to Home Screen”

On Content Creators’ Duty to Disclose AI Generated Content

Political Theory
Freedom
Social Media
Normative Theory
William Chan
University of Cambridge
William Chan
University of Cambridge

Abstract

Social media platforms are flooded with AI-generated content. Many creators of Instagram reels or TikTok videos, for example, use AI to create animations, voiceovers, pictures, and so on. Some platforms require users to disclose that their content is generated by AI; some have proposed that existing regulatory frameworks (e.g. GDPR, EU AI Act) for AI governance imply a legal duty of disclosure on content creators utilising generative AI tools; and some are proposing new legislation on mandatory disclosure. Together, these considerations highlight the normative and practical significance of the grounds for content creators’ duty, if any, to disclose AI-generated content. My aim is to explore two major arguments for such a duty: (1) the Argument from Regulatory Compliance and (2) the Argument from Risk Avoidance. According to the Argument from Regulatory Compliance, existing regulatory frameworks for AI and data impose transparency, explainability, accountability, and copyright obligations on users, which imply content creators’ duty to disclose AI-generated content on social media platforms. The Argument from Risk Avoidance, in contrast, is consequentialist and addresses the specific risks of content creators not disclosing their use of AI. Some of these risks involve violations of stakeholders’ interests (e.g. content creators, platforms, viewers), whereas others concern the harmful social consequences of the presence of undisclosed AI-generated content in the online world. After evaluating the strengths and weaknesses of both arguments, I will discuss an additional dimension of content creators’ duty of AI disclosure that is often overlooked: how exactly they should disclose their use of generative AI. I argue that, without clarifying this question, the key considerations underpinning the Arguments from Regulatory Compliance and Risk Avoidance will not be adequately addressed whatsoever.