"Unlimited Realm of Exploration and Experimentation": Methods and Motivations of AI-Generated Sexual Content Creators

AI-generated media is radically changing the way content is both consumed and produced on the internet, and in no place is this potentially more visible than in sexual content. AI-generated sexual content (AIG-SC) is increasingly enabled by an ecosystem of individual AI developers, specialized third-party applications, and foundation model providers. AIG-SC raises a number of concerns from old debates about the line between pornography and obscenity, to newer debates about fair use and labor displacement (in this case, of sex workers), and spurred new regulations to curb the spread of non-consensual intimate imagery (NCII) created using the same technology used to create AIG-SC. However, despite the growing prevalence of AIG-SC, little is known about its creators, their motivations, and what types of content they produce. To inform effective governance in this space, we perform an in-depth study to understand what AIG-SC creators make, along with how and why they make it. Interviews of 28 AIG-SC creators, ranging from hobbyists to entrepreneurs to those who moderate communities of hundreds of thousands of other creators, reveal a wide spectrum of motivations, including sexual exploration, creative expression, technical experimentation, and in a handful of cases, the creation of NCII.

💡 Research Summary

The paper “Unlimited Realm of Exploration and Experimentation: Methods and Motivations of AI‑Generated Sexual Content Creators” offers the first systematic, qualitative investigation of the people who produce AI‑generated sexual content (AIG‑SC). Recognizing that generative AI has democratized the creation of erotic images, text, audio, and video while simultaneously enabling the rapid production of non‑consensual intimate imagery (AIG‑NCII) and child sexual abuse material (AIG‑CSAM), the authors set out to fill a gap in the literature that has focused almost exclusively on the abusive side. They pose four research questions: (1) who creates AIG‑SC and how they become involved; (2) what types of AIG‑SC are produced and by what technical means; (3) why creators generate this content; and (4) how perpetrators of AIG‑NCII operate and what motivates them.

Methodologically, the study conducts semi‑structured interviews with 28 participants representing a spectrum from hobbyist creators, “jailbreakers” who subvert model safeguards, sex workers, tool developers, to moderators of large creator communities. The recruitment deliberately targets communities that publicly prohibit non‑consensual content, yet three interviewees disclosed that they had produced AIG‑NCII, and a fourth reported a friend who had done so. Ethical safeguards were applied throughout, including gender‑matched interviewers and consultation with scholars experienced in interviewing perpetrators of sexual violence.

The findings are organized around the four research questions. First, most creators have backgrounds in computing, the arts, or the sex industry, and they entered AIG‑SC through a variety of triggers: encountering safety filters on general‑purpose models, curiosity about a niche fetish, desire to learn AI skills, or the appeal of a supportive community. Second, the content is dominated by text and images, with emerging interest in audio and future video generation. Three production modalities emerge: (1) customized pipelines that combine prompt engineering, low‑rank adaptation (LoRA) fine‑tuning, scripting (often Python), and post‑processing tools such as Photoshop; (2) UI‑bound approaches that rely on direct prompting of specialized services (e.g., character.ai) or general models, sometimes using jailbreak techniques to bypass restrictions; and (3) commissions, where creators outsource or collaborate on content creation. Third, motivations are multifaceted: satisfying specific sexual interests or fetishes, expressing oneself creatively, challenging technical limits and acquiring new skills, earning money or gaining status within a community, and resisting perceived sexual censorship. The community functions as a learning hub, with technically proficient members mentoring newcomers. Fourth, the three participants who admitted to creating AIG‑NCII employed the same technical toolbox as AIG‑SC creators—face‑swapping, fine‑tuning, and jailbreaks—but applied it to images of acquaintances, celebrities, or strangers. Their self‑justifications ranged from “no real harm” to framing the activity as a security‑research exercise, while also citing sexual gratification, boredom, community prestige, and financial reward.

The discussion highlights a policy tension: broad platform bans on sexual AI content risk suppressing legitimate expression and the livelihoods of sex workers, whereas targeted interventions against non‑consensual content are technically challenging but essential. The authors advocate for nuanced governance that combines precise detection of AIG‑NCII, community‑driven moderation, and platform policies that protect free speech while preventing abuse. They also note the potential for AI to empower sex workers economically, urging safeguards against labor displacement. Limitations include a sample skewed toward English‑speaking online spaces and reliance on self‑reporting, which may introduce social desirability bias.

In conclusion, the study maps the ecosystem of AIG‑SC creators, detailing their backgrounds, technical practices, and motivations, and exposing the overlap with non‑consensual content production. These insights provide empirical grounding for legislators, platform operators, and community leaders seeking to craft balanced regulations that curb abuse without stifling legitimate creative and sexual expression in the age of generative AI.