OpenAI Is Looking At How To ‘Responsibly’ Make NSFW AI-Generated Material NSFW
Illustration 286279026 © Anastasiya Lavrik | Dreamstime.com
OpenAI is pretty buttoned-up when it comes to Not Safe for Work (NSFW) content, blocking attempts to generate sexually explicit material on its ChatGPT and DALL-E 3 tools. However, with the rapid evolution of artificial intelligence, it appears to be growing up too. The AI leader is contemplating revising its policy to allow NSFW erotica under specific, age-appropriate conditions.
As of now, OpenAI’s guidelines, which were last updated on January 10, 2024, prioritize safety, responsibility, and user control. These rules outrightly ban the use of its technology to support illegal activities, cause harm, or spread damaging material, including any sexually explicit or suggestive content.
However, its newly released “model spec,” a kind of draft for how the company’s AI should operate, hints at possible changes. OpenAI is now questioning whether it can “responsibly” enable the generation of adult content in a controlled manner.
Rather than being the one to create AI pornography, the organization is kicking off discussions on whether and how erotic text and imagery could be integrated into its offerings without breaching ethical standards. This exploration is part of a broader effort to understand user and societal expectations regarding model behavior in sensitive areas.
OpenAI’s re-evaluation of its stance on NSFW content is happening against a backdrop of increasing use of AI to create deepfake pornography and synthetic nudes. To keep its technology safe and reliable, the Microsoft-backed firm trains its models to reject harmful instructions and continuously updates them to reduce the likelihood of generating damaging content.
[via NPR and Futurism, cover illustration 286279026 © Anastasiya Lavrik | Dreamstime.com]
