Experiences with AI-Generated Pornography: A Quantitative Content Analysis of Reddit Posts.

Résumé

With the rise of generative artificial intelligence (AI), a new form of pornography has emerged: AI-generated pornography (AI pornography, AI adult content, synthetic pornography). This phenomenon encompasses both entirely fictional content and deepfake pornography, which involves non-consensual depictions of celebrities and private individuals. While non-consensual deepfake pornography has sparked widespread concern, the broader range of experiences associated with AI pornography remains under-researched. This study aimed to address this gap by examining the spectrum of AI pornography experiences as shared on Reddit, adopting a communication process perspective. Specifically, it explored the following research questions (RQ): How do Reddit users address the production (RQ1), content (RQ2), use (RQ3), effects (RQ4), and ethical-legal implications (RQ5) of AI pornography in their posts? A total of 390 English-language public Reddit posts related to AI pornography were collected in August 2024 and analyzed using a manual quantitative content analysis guided by a reliability-tested codebook (mean Cohen's kappa = 0.88). The analysis revealed that the production (RQ1: 59.5%) and content (RQ2: 60.8%) of AI pornography were the most frequently discussed topics, while use experiences (RQ3: 12.8%) were the least represented. Effects (RQ4: 37.2%) and ethical-legal implications (RQ5: 35.1%) were addressed in approximately one-third of posts. The AI pornography experiences addressed in Reddit posts ranged from outrage and concern regarding its actual and potential harms to curiosity, enjoyment, and economic benefits, underscoring the need for a nuanced response from policymakers, technology developers, educators, and mental health professionals.