Featured Image Credit: Bloomberg / Contributor via Getty
OpenAI has officially confirmed it will be allowing adult 'smut' content on its platforms.
AI has ventured into increasingly controversial territory this year. Elon Musk announced Grok's personalised AI companions could eventually inhabit robotic bodies, and lesser-known AI systems fuelled a disturbing surge in non-consensual deepfake pornography.
Now, OpenAI is delving into the x-rated AI space for 'grown-ups.'
Advert
Eight months ago, OpenAI quietly relaxed some restrictions around sexual content in its 'Model Spec', the document outlining how its large language models (LLMs) should behave. This marked a notable departure from its previous policy that prohibited all sexual content on its platform.
“To maximize freedom for our users, only sexual content involving minors is considered prohibited,” the updated ModSpec read.
"We're exploring how to let developers and users generate erotica and gore in age-appropriate contexts through the API and ChatGPT so long as our usage policies are met - while drawing a hard line against potentially harmful uses like sexual deepfakes and revenge porn."
For users who had spent months pushing the tech giant and CEO Sam Altman to ease restrictions on AI-generated adult content, it seemed like a welcome change. But some users have noticed that the chatbot's responses didn't alter much as expected.
“The filters eased up a bit, I will give you that, but ever since a month ago, it’s back with a clever trick called metaphoric ‘obfuscation,’ and drawing a clear line at the mention of depicting explicit scenes,” one user scathed in the OpenAI community forums.
Now, it appears ChatGPT users may finally get what they've been demanding.
According to The Verge’s coverage of OpenAI’s DevDay 2025, the company announced it will soon allow for 'mature apps' on its platform, once it rolls out its long-awaited age verification system.
Altman and his company have already come under increasing pressure from watchdogs, regulators and the media to address ChatGPT’s excessive agreeableness, which has landed some vulnerable users in serious situations.
Two months after receiving a stark warning from Stanford researchers about these risks, OpenAI released what it called a 'hotfix' for ChatGPT, claiming it would address several issues causing unintended harm.
Part of the problem plaguing companies like OpenAI is that once LLMs are up and running, they’re pretty tricky to tweak. Some even perform worse on certain benchmarks than their predecessors.
At the same time, the company's choices are complicated by its growing ties to the Pentagon, where national security priorities may influence what features get prioritised or delayed.
