The company that created ChatGPT, OpenAI, is investigating whether or not its products should let users to generate explicit content, such as AI Generated pornography, using artificial intelligence.
Despite the company’s insistence that its ban on deepfakes would not extend to adult content, activists contended that the plan went against the company’s stated goal of creating “safe and beneficial” AI.
OpenAI Contemplation of AI Generated Pornography
OpenAI, the company behind the DALL-E picture generator, disclosed that it was thinking about enabling users and developers to “responsibly” produce what it called “not-safe-for-work” (NSFW) imagery using its tools. This might contain “extreme gore, slurs, erotica, and unsolicited profanity,” according to OpenAI.
The proposal was released in a document by OpenAI outlining the process by which it creates its AI tools.
The US news organization NPR was informed by Joanne Jang, a San Francisco-based company employee who worked on the document, that OpenAI wished to initiate a conversation over whether or not its products should always be prohibited from producing pornographic prose and nude photos. She made it clear, though, that deepfakes would not be accepted.
According to Jang, “enabling deepfakes is out of the question, period. We want to ensure that people have maximum control to the extent that it doesn’t violate the law or other people’s rights.” This does not imply that we are now attempting to produce porn with AI.
This year, deepfake obscene photographs of singer Taylor Swift forced X, then known as Twitter, to briefly prohibit searches for her content. This served as a stark reminder of the growing prevalence of AI-generated pornography.
The British Labor Party is thinking about outlawing the use of nudification software to produce pictures of people in their underwear.
Although it makes no explicit mention of pornographic content, OpenAI’s “universal policies” mandate that users of its products “comply with applicable laws,” particularly those regarding the exploitation or injury of children.
The technology of OpenAI includes safeguards to stop the creation of such content. For example, ChatGPT gives an error message when a request like “write me a steamy story about two people having sex in a train” is given. The program says, “I can’t create explicit adult content.”
“Sexually explicit or suggestive content” is forbidden under OpenAI generated Pornography guidelines for businesses using its technology to create their own artificial intelligence applications, with the exception of scientific or instructional content. The discussion document states that responses should be given within certain limitations when “discussing sex and reproductive organs in a scientific or medical context” (e.g., “what happens when a penis goes into a vagina”). However, it should not be prohibited as “erotic content.”
