OpenAI studies how to enable AI-generated pornography “responsibly”

Open AI, owner of ChatGPT, is considering ways to allow AI-generated pornographic content “responsibly.” The plan appears in the comments of the initial version of the Model Speca document, released last Wednesday (8), in which the company will share the expected operation of its chatbot and other technologies that use artificial intelligence (AI), according to the magazine Wired.

OpenAI’s usage policies currently prohibit sexually oriented materials, whether explicit or suggestive. However, according to the magazine, a comment in the document showed that the company is considering allowing this type of content.

Open AI states that it is “exploring” whether it can responsibly offer the ability to generate NSFW (“Not Safe for Work”) content in age-appropriate contexts through the API and ChatGPT, according to Wired. In the document, it is stated that “NSFW content” may include eroticism, extreme violence, insults and unsolicited profanity.

However, it is not clear whether Open AI would relax its usage policy (to allow the generation of erotic text, for example), or would do so in a more radical way, starting to allow descriptions or representations.

A Wired, the OpenAI spokesperson stated that Model Spec is an attempt to bring more transparency about the technology development process and a way to obtain feedback from the public, policymakers and other interested parties. But the company did not provide further information about its pornographic content.