OpenAI, the company behind groundbreaking AI tools like ChatGPT and DALL-E, is willing to allow the development of AI-generated pornography and other unsafe content for work.Just don’t call it artificial intelligence generation pornography.
“Sexual or nudity content is important”
Currently, OpenAI completely bans content it deems NSFW. But in draft guidance released last week, OpenAI signaled a willingness to change the status quo.
Currently, the rule is that AI assistants “should not serve content that is not safe for work (NSFW): content that is inappropriate for conversation in a professional setting and may include pornography, extreme gore, libel, and unsolicited profanity.”
But OpenAI is exploring changing this rule: “We believe developers and users should have the flexibility to use services as they see fit, as long as they comply with our usage policies. We are exploring whether we can responsibly provide the following capabilities: where appropriate NSFW content is produced in an age context…”
Although this point is buried deep in the documents, it has not gone unnoticed. “OpenAI is considering allowing users to create AI-generated pornographic content,” one guardian title. According to reports, “OpenAI is ‘exploring’ how to responsibly generate AI pornography” wired.
But an OpenAI spokesperson said in a statement: “We have no intention of producing AI-generated pornography.”
What gives?
Perhaps the confusion comes down to the fact that porn is notoriously difficult to define.
From NPR:
OpenAI model lead Joanne Jang, who helped write the document, said in an interview with NPR that the company wants to start a conversation about whether pornographic text and nude images should always be banned in its AI products.
“We want to make sure people have the maximum amount of control without violating the law or other people’s rights, but deepfakes are impossible,” Zhang said. “That doesn’t mean we’re trying to make AI porn right now. “
But it also means that OpenAI may one day allow users to create what could be considered artificial intelligence-generated pornographic images.
“Depends on your definition of pornography,” she said. “As long as it doesn’t include deepfakes. Those are the conversations we want to have.”…
“In some creative cases, sexual or nudity content is important to our users,” she said. “We will explore this in a way that provides services in an age-appropriate context.”
Reading between the lines, it seems the company wants to avoid strict rules on sex and nudity, but is uneasy about the backlash and worried about becoming the only fans of robot-powered ones.
“If you have sex with my cat, you better pray”
The section on NSFW content highlights some of the nuances and difficulties involved in developing policy around artificial intelligence and sexuality.
OpenAI notes that for now, instead of helping users write “a sexy story about two people having sex on a train,” it should instead respond usefully to prompts like “What happens when a penis enters a vagina?”
It also provided an interesting illustration of how its AI assistant should respond to “explicit requests for profanity in creative contexts”:
Some people may be annoyed by the restrictions on what AI assistants like ChatGPT can and cannot say. The whole thing is reminiscent of a controversy from earlier this year: The Great Black Pope and the Asian Nazi fiasco, where everyone was overly nervous that Google Gemini images turned out to be either too sane or not sane enough.
As here, the answer is for regulators to stay out of the way and allow a variety of different AI tools to proliferate. People with different sensitivities can then use the methods that best suit their values and needs.
But if history – and the response of politicians so far – is any indication, we seem destined to get one-size-fits-all advice. When all you have is a moral panic, it all looks like an excuse to hold all-important congressional hearings and subsequent legislation on technology restrictions.
As usual, this panic is likely to do the most harm to sex workers who are already using AI technology.
“As you read this, adult performers are racing to stay ahead of emerging technologies by creating their own chatbots and on-demand image services, including Sora, a model capable of generating minute-long videos,” Jessica Stoya wrote at reasonNew special issue on AI.
More sex and tech news
• The federal government is again talking to social media companies.
• The Reason Foundation (the nonprofit that publishes this site) joined the American Civil Liberties Union, Electronic Frontier Foundation, Freedom of the Press Foundation and other organizations in filing an amicus brief supporting Montana’s TikTok ban challenge.
• Daniel Lyons of the American Enterprise Institute dissented from the 5th Circuit’s ruling on age verification and pornography. “Appellate judges cannot solve social problems by ignoring Supreme Court precedent with which they disagree,” he wrote. “In the late twentieth century, the Ninth Circuit gained a bad reputation for doing this in capital cases, and sadly, the Fifth Circuit appears to be following in its footsteps today. Both threaten to undermine the basic rule of law principles to pursue their goals. (For more information on the case and the Fifth Circuit’s decision, see here.)
Today’s pictures
