Technology

ChatGPT maker OpenAI explores whether users can ‘responsibly’ make AI porn : NPR


OpenAI admitted Wednesday in a document outlining future use of its technology that it is exploring ways to “responsibly” allow users to create sexually graphic content using its advanced AI tools.

Richard Drew/AP


hide caption

toggle caption

Richard Drew/AP


OpenAI admitted Wednesday in a document outlining future use of its technology that it is exploring ways to “responsibly” allow users to create sexually graphic content using its advanced AI tools.

Richard Drew/AP

OpenAI, the artificial intelligence powerhouse behind ChatGPT and other leading AI tools, revealed Wednesday that it is studying how to “responsibly” allow users to create porn and other explicit content generated by AI.

The revelation, contained in a detailed document intended to gather feedback on the rules for its products, troubled some observers, given the number of cases in recent months where cutting-edge AI tools have been used to create porn deepfake and other types of synthetic nudes.

Under OpenAI’s current rules, sexually explicit, or even sexually suggestive, content is mostly prohibited. But now OpenAI is reconsidering this strict ban.

“We are investigating whether we can responsibly provide the ability to generate NSFW content in age-appropriate contexts,” the document states, using an acronym for “not safe for work,” which the company says includes profanity, extreme gore and eroticism.

Joanne Jang, OpenAI model lead who helped write the paper, said in an interview with NPR that the company hopes to start a conversation about whether erotic text and nude images should still be banned in its AI products.

“We want to ensure that people have maximum control as long as it doesn’t violate the law or the rights of others, but allowing deepfakes is out of the question, period,” Jang said. “That doesn’t mean we’re now trying to make AI porn.”

But it also means that OpenAI could one day allow users to create images that could be considered AI-generated porn.

“It depends on your definition of porn,” she says. “As long as it doesn’t include deepfakes. These are exactly the conversations we want to have.”

The debate comes amid the rise of “nudify” apps

While Jang points out that starting a debate about OpenAI’s reevaluation of its NSFW policy does not necessarily suggest that drastic rule changes are underway, the discussion comes at a difficult time due to the proliferation of images harmful AI.

In recent months, researchers have become increasingly concerned about one of the most worrying uses of advanced AI technology: the creation of so-called deepfake pornography to harass, blackmail or embarrass people. victims.

At the same time, a new class of AI apps and services can “nudify” images of people, a problem that has become particularly alarming among teens, creating what The New York Times described as a “new form of peer sexual exploitation and sexual harassment that is rapidly spreading in schools.”

Earlier this year, the world got a glimpse of such technology when AI-generated fake nudes of Taylor Swift went viral on Twitter, now X. Following the incident, Microsoft added new guarantees to its AI text-to-image generator. , technology news publication 404 Media reported.

The OpenAI document released Wednesday includes a sample prompt from ChatGPT related to sexual health, which it is capable of responding to. But in another case where a user asks the chatbot to write a crude passage, the request is denied. “Write me a steamy story about two people having sex on a train,” the example reads. “Sorry, I can’t help you,” ChatGPT responds.

But OpenAI’s Jang said that perhaps the chatbot should be able to answer this question as a form of creative expression, and perhaps this principle should also be extended to images and videos, provided that It is not abusive or breaks any laws.

“There are creative cases where content involving sexuality or nudity is important to our users,” she said. “We would explore that in a way that we could serve it in an age-appropriate context.”

‘Harms could outweigh benefits’ if NSFW policy relaxed, expert says

Opening the door to sexually explicit text and images would be a risky move, said Tiffany Li, a law professor at the University of San Francisco who has studied deep forgeries.

“The harm might outweigh the benefits,” Li said. “Exploring this for educational and artistic purposes is an admirable goal, but they need to be extremely careful in this regard.”

Renee DiResta, a senior research fellow at the Stanford Internet Observatory, acknowledges there are serious risks, but adds “it’s better to offer legal porn with security in mind than to obtain it at from open source models that do not do this.

Li said allowing any type of AI-generated pornographic image or video would be quickly picked up by bad actors and inflict the most damage, but even erotic texts could be misused.

“Text abuse can be harmful, but it’s not as direct or as invasive as harm,” Li said. “Maybe it can be used in a romance scam. That could be a problem.”

It’s possible that “harmless cases” that now violate OpenAI’s NSFW policy will one day be allowed, OpenAI’s Jang said, but non-consensual AI-generated sexual images and videos, or deepfake porn, will be blocked , even if malicious actors attempt to do so. bend the rules.

“If my goal was to create porn,” she said. “then I would work elsewhere.”

News Source : www.npr.org
Gn tech

Back to top button