OpenAI and the Development of Adult-Oriented Material
On May 8, 2024, the pioneering AI organization OpenAI, well-known for developing the advanced AI platform commonly known as ChatGPT, unveiled preliminary protocols for potentially enabling the production of adult-oriented material, including erotic content, as WIRED has reported.
The preliminary framework, titled Model Spec, reflects OpenAI’s progression from its prior policy of disallowing explicit content, taking into account the societal and ethical ramifications of potentially sanctioning the generation of adult content. The Model Spec includes a note stating, “We are considering the possibility of responsibly offering the capacity to craft adult content in contexts that are suitable for the audience’s age via the API and ChatGPT.” OpenAI is proactively engaging in understanding the public and user expectations concerning the comportment of its models within this sensitive domain.
The debate regarding synthetic adult content generated by AI, such as pornographic materials, has garnered significant attention at OpenAI. Despite assertions made by OpenAI’s representative, Niko Felix, who has stated, “our models are not designed to generate AI-based erotica,” the organization acknowledges the user-driven divergence in defining adult material. Joanne Jang, a contributor at OpenAI, has underscored the ambiguity to NPR, asserting, “It’s all about how you define pornography.”
Navigating the Moral Maze and Contributing to Public Debates
In light of growing apprehensions over the possible misapplication of AI tools, such as deepfake pornographic content that can lead to harassment and breaches of privacy, OpenAI has sought to underscore its commitment to transparency and community involvement. Grace McGuire, another OpenAI representative, spoke to WIRED about the intentions behind the Model Spec, explaining the document’s purpose is to “foster greater openness concerning our development procedures and to solicit a broad array of insights and opinions from stakeholders across society, including the public, policymakers, and industry observers.”
Danielle Keats Citron, a law professor at the University of Virginia, has voiced concerns over the negative impacts of deepfake pornography and unauthorized intimate synthesized imagery, pointing out potential repercussions on employment, personal safety, and freedom of expression.
While OpenAI’s service parameters strictly prohibit impersonation without consent, precluding unsanctioned explicit imagery, questions about moderation and enforcement of these rules persist in discussions about how to prevent misuses of the technology. As OpenAI confronts these issues, it is reminiscent of the challenges Microsoft faced with regulating its generative AI tool after reports of misuse emerged from 404 Media.
As discussions continue, industry observers and the general public are poised to closely monitor further communications from OpenAI, particularly in relation to their oversight of explicit content generation. The organization faces the daunting task of steering through the intricate interplay of AI technology with the production of adult content and the attendant societal implications..
