In a pivotal move by OpenAI, the innovators behind the groundbreaking ChatGPT, the exploration into the production of adult content through AI is taking shape.
On a recent Wednesday, OpenAI initiated draft guidelines that aim to define the conduct for ChatGPT and its AI counterparts. Within the detailed Model Spec document, discussions have emerged about the potential foray into the sexuality explicit sector—currently at odds with present restrictions against such content.
The organization is considering the controlled generation of Not Safe For Work material, tapping into its API and the capabilities of ChatGPT itself. Commentary within the Model Spec mentions OpenAI’s goal to align with the expectations of users and the broader society regarding model behavior in this sensitive arena. The range of NSFW content under review encompasses everything from eroticism to harsher forms of graphic imagery, hate speech, and unexpected vulgarity. The degree to which OpenAI may relax its policy to embrace such content integration remains undefined.
Addressing Complexities and Operating Transparently
OpenAI’s representative, Niko Felix, informed WIRED that the company’s models are not intended to create AI-derived pornography. That said, Joanne Jang from OpenAI points out that ultimately, it will be up to users to determine if the technology will be applied to adult content; the notion of pornography can be subjective.
Grace McGuire, another spokesperson for OpenAI, underscored the commitment to transparent communication of its developmental efforts with the Model Spec acting as a vehicle for gathering diverse perspectives, including those of the general public, policy designers, and stakeholders. However, the specifics of its investigative processes and the community feedback received regarding this issue were not disclosed by the entity.
The creation of pornography through artificial intelligence, particularly with the ever-expanding presence of deepfake technologies, has become a burgeoning concern, with its utilization for harassment, notably targeting women and minors without consent. Legal and ethical quandaries loom large, with scholars such as Danielle Keats Citron from the University of Virginia School of Law highlighting the troubling interest in AI-enabled NSFW content, stressing the ubiquitous and destructive nature of such infringements.
Despite the challenging path to responsibly crafted AI-generated NSFW content, riddled with regulatory and ethical hurdles, OpenAI’s existing stance against the creation of involuntary explicit material holds steadfast. Yet, the potential for abuse of the technology underscores concerns about the effectiveness of moderation to avert misuse. In one reported incident of abuse that generated explicit imagery of celeb Taylor Swift, adjustments were made to a Microsoft AI tool to thwart future similar transgressions.
The evolution of AI and its associated challenges, especially in maintaining ethical frameworks in tandem with technological growth, continues to provoke industry-wide debate. OpenAI’s ultimate decisions and forthcoming policies concerning AI-rendered pornography remain, for the most part, undisclosed.
