OpenAI is not only interested in us generating “spicy content”, but they also want us to do it correctly

OpenAI unveiled a draft of its documentation called Model Spec this Wednesday, outlining the company’s intention to regulate the behavior of its artificial intelligences, including ChatGPT. Among the notable points is OpenAI’s consideration to regulate the production of pornographic content and other explicit materials, as their current policies prohibit such content.

ChatGPT DOWNLOAD

According to the document, OpenAI is exploring the possibility of allowing the generation of NSFW content (an acronym for “not safe for work”) through its API and ChatGPT. Among the contents classified as NSFW in the document are erotica, extreme gore, unsolicited insults, and blasphemies.

Grace McGuire, spokesperson for OpenAI, stated to WIRED that Model Spec aims to bring transparency to the development process and obtain feedback from various stakeholders. However, she refused to provide specific details about the exploration of generating explicit content by the company.

The possibility of OpenAI allowing explicit content has generated a lot of uncertainty, especially in the context of the rise of AI-generated pornography, such as deepfakes. These explicit videos or images, created with AI, depict porn actors and actresses whose faces have been replaced with real people without their consent, which can lead to situations of harassment and abuse.

Danielle Keats Citron, a professor at the University of Virginia School of Law, has expressed concern about this direction, pointing out the serious impacts that AI-generated explicit content can have on the privacy and security of affected individuals.

Although OpenAI’s policies prohibit identity impersonation without permission, the potential regulation for generating NSFW content raises many questions about the company’s ability to effectively moderate the malicious use of its tools. Microsoft had to deal with a similar problem, making a series of changes to its policies after it was confirmed that its generative AI had been used to create deepfake images of Taylor Swift that were distributed on Twitter.

ChatGPT DOWNLOAD