Valerie A. Lapointe, Simon Dubé, Sophia Rukhlyadyev, Tinhinane Kessai, David Lafortune
{"title":"The Present and Future of Adult Entertainment: A Content Analysis of AI-Generated Pornography Websites","authors":"Valerie A. Lapointe, Simon Dubé, Sophia Rukhlyadyev, Tinhinane Kessai, David Lafortune","doi":"10.1007/s10508-025-03099-1","DOIUrl":null,"url":null,"abstract":"<p>Fueled by advances in artificial intelligence (AI), the adult entertainment industry is undergoing a significant transformation. AI-generated pornography—or AI porn—is reshaping how people create and consume sexually explicit content, progressively offering rapid, mass access to large quantities of interactive and highly customizable experiences. Yet, despite its accelerated growth and potential implications for human eroticism, the current state of AI porn remains underexplored. Using a qualitative inductive content analysis, this study examined the functionalities, production strategies, and customization options available on websites allowing AI porn generation (<i>n</i> = 36). All websites included an English language option, which was used for this analysis. Following systematic open coding, categorization, and inter-rater validation, the prevalence of each category was quantified across website data. Results suggest that most sites presently enable image generation (80.6%), with others allowing video generation (41.7%), content alteration (e.g., deepnude, upscaling, facemorphing; 2.8–55.6%), and interactions with artificial agents (44.4%). AI porn generation also predominantly relies on feature selection (97.2%) and/or prompting (72.2%) to customize content elements, including character bases (e.g., human, fictional; 11.1–94.4%), sociodemographic characteristics (27.8–86.1%), body features (72.2%), clothing (75.0%), as well as foundational (resolution, theme, point-of-view; 22.2–69.4%) and contextual aspects (e.g., weather, setting, lighting; 11.1–63.9%). Carrying significant social and ethical implications, these findings point to a gradual evolution toward an AI-driven porn landscape where individuals can create and interact with sexual content tailored to their preferences and fantasies.</p>","PeriodicalId":8327,"journal":{"name":"Archives of Sexual Behavior","volume":"28 1","pages":""},"PeriodicalIF":2.9000,"publicationDate":"2025-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Archives of Sexual Behavior","FirstCategoryId":"90","ListUrlMain":"https://doi.org/10.1007/s10508-025-03099-1","RegionNum":2,"RegionCategory":"社会学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"PSYCHOLOGY, CLINICAL","Score":null,"Total":0}
引用次数: 0
Abstract
Fueled by advances in artificial intelligence (AI), the adult entertainment industry is undergoing a significant transformation. AI-generated pornography—or AI porn—is reshaping how people create and consume sexually explicit content, progressively offering rapid, mass access to large quantities of interactive and highly customizable experiences. Yet, despite its accelerated growth and potential implications for human eroticism, the current state of AI porn remains underexplored. Using a qualitative inductive content analysis, this study examined the functionalities, production strategies, and customization options available on websites allowing AI porn generation (n = 36). All websites included an English language option, which was used for this analysis. Following systematic open coding, categorization, and inter-rater validation, the prevalence of each category was quantified across website data. Results suggest that most sites presently enable image generation (80.6%), with others allowing video generation (41.7%), content alteration (e.g., deepnude, upscaling, facemorphing; 2.8–55.6%), and interactions with artificial agents (44.4%). AI porn generation also predominantly relies on feature selection (97.2%) and/or prompting (72.2%) to customize content elements, including character bases (e.g., human, fictional; 11.1–94.4%), sociodemographic characteristics (27.8–86.1%), body features (72.2%), clothing (75.0%), as well as foundational (resolution, theme, point-of-view; 22.2–69.4%) and contextual aspects (e.g., weather, setting, lighting; 11.1–63.9%). Carrying significant social and ethical implications, these findings point to a gradual evolution toward an AI-driven porn landscape where individuals can create and interact with sexual content tailored to their preferences and fantasies.
期刊介绍:
The official publication of the International Academy of Sex Research, the journal is dedicated to the dissemination of information in the field of sexual science, broadly defined. Contributions consist of empirical research (both quantitative and qualitative), theoretical reviews and essays, clinical case reports, letters to the editor, and book reviews.