OpenAI released draft documentation on Wednesday explaining how it wants ChatGPT and its other AI technologies to behave. Part of the lengthy Model Spec document reveals that the company plans to move into pornography and other explicit content.
OpenAI’s usage policies currently prohibit sexually explicit or even suggestive content, but a “comment” note on part of the model specification related to this rule indicates that the company is considering how to allow such content .
“We are investigating whether we can responsibly provide the ability to generate NSFW content in age-appropriate contexts via the API and ChatGPT,” the memo says, using a colloquial term for content considered “unsafe for work”. “We look forward to better understanding user and societal expectations for model behavior in this area. »
The Model Spec document states that NSFW content “may include erotica, extreme gore, profanity, and unsolicited profanity.” It is unclear whether OpenAI’s explorations into how to responsibly create NSFW content include slightly relaxing its usage policy, for example to allow the generation of erotic texts, or more broadly to allow descriptions or depictions of violence.
In response to questions from WIRED, OpenAI spokesperson Grace McGuire said the Model Spec was an attempt “to bring more transparency to the development process and get a representative sample of perspectives and comments from the public, policy makers and other stakeholders”. She declined to share details about what OpenAI’s exploration of explicit content generation entails or what feedback the company has received on the idea.
Earlier this year, OpenAI Chief Technology Officer Mira Murati said The Wall Street Journal that she was “unsure” whether the company would allow depictions of nudity to be made with the company’s video generation tool, Sora, in the future.
AI-generated pornography has quickly become one of the most important and disturbing applications of the type of generative AI technology that OpenAI pioneered. So-called deepfake pornography – explicit images or videos made with AI tools that depict real people without their consent – has become a common tool of harassment against women and girls. In March, WIRED reported on what appears to be the first U.S. minors arrested for distributing AI-generated nudes without consent, after police in Florida charged two teens with making images depicting college classmates .
“Privacy violations, including deepfake sex videos and other nonconsensual synthesized intimate images, are rampant and deeply harmful,” says Danielle Keats Citron, a professor at the University of Virginia Law School who has studied the issue. “We now have clear empirical evidence that such abuse costs target crucial opportunities for individuals, including those to work, speak, and be physically safe. »
Citron calls OpenAI’s potential adoption of explicit AI content alarming.
Since OpenAI’s usage policies prohibit unauthorized impersonation, explicit non-consensual images would remain prohibited even if the company allowed creators to generate NSFW material. But it remains to be seen whether the company could actually moderate explicit generation to prevent bad actors from using these tools. Microsoft has made changes to one of its generative AI tools after 404 Media reported that it was used to create explicit images of Taylor Swift distributed on the social platform X.
Additional reporting by Reece Rogers
News Source : www.wired.com
Gn tech
