Pornography and Erotica Made the Web and Now ChatGPT is Joining the Adult Entertainment World

Don’t look now, but generative artificial intelligence (GenAI) is already in the pornography business as AI porn is entering the mainstream. So, it shouldn’t be too surprising to read that the CEO of OpenAI, Sam Altman, wants ChatGPT to treat adult users like adults, which means including using the technology to create and view erotica. Altman’s ChatGPT joins Elon Musk’s Grok and Mark Zuckerberg’s Meta AI in the pursuit of creating virtual assistants that serve as friends with benefits.

Pornography has played a significant role in the growth of the Internet from the dawn of the World Wide Web to today. In the 1990s, it was pornography that propelled advances in web-delivered images, video, and interactive consumer experiences. The pornography industry was among the first to do web advertising. The industry was on the ground floor of pop ads, the first pay-per-view pay walls, and subscription options.

In a July 3, 1995, Time magazine cover story, entitled “On Screen Near You: Cyberporn,” it suggested that the pervasiveness and popularity of pornography was dominating web traffic, ringing alarm bells in Congress and leading to 1996 legislation to stop cybererotica. Almost thirty years later, an MIT Technology Review article appearing on August 26, 2024, described a new evolution in web pornography powered by AI. This article stated that AI porn needed a national conversation in the United States and other countries. Was AI porn about freedom of speech, or about morality and criminality?

There is a dark side to AI-generated sexualized images. They can be created by manipulating real ones, such as replacing in a video, the face of a “porn star” with one of the girl next door. The article noted that using an AI porn generator, a user could “check boxes on a list of options as long as the Cheesecake Factory menu to create their ideal scenes: categories like male, female, and trans; ages from 18 to 90; breast and penis size; details like tan lines and underwear color; backdrops like grocery stores, churches, the Eiffel Tower, and Stonehenge; even weather, like tornadoes.”

The Altman comment stating that ChatGPT will offer subscribers the ability to customize its AI for personal use leaves many new options. Subscribers can give their AI a distinctive personality. They could introduce it to digitized versions of old home movies and photos, turning old 8-millimetre film reels stored in the basement or attic into the latest “blue movies.” If you don’t know the blue movie reference, ask an AI. It knows.

Altman’s age-gated ChatGPT for adult subscribers could become a virtual boyfriend or girlfriend that is subject to the whim of its user and capable of doing almost anything imaginable, including virtual sex. To describe this new capability in a nutshell, it represents an industry paradigm shift with enormous moral and safety implications.

Remember, we already have an Internet awash in sexualized content. Some, therefore, may see this new AI capacity as no different from the mid-1990s when traffic on pornography and erotica websites became significant. But critics see in AI-porn risk, the exploitation and sexualization of the vulnerable and, in particular, children.

Can regulating AI-generated content on the Internet work without freedom of speech and thought threatened? Is the issue of AI porn no different than any other related to this emerging technology?

Since ChatGPT 3.0 emerged three years ago, many, including Geoffrey Hinton, the so-called father of AI, have called for the establishment of guardrails. Some have answered the call.

In 2025, the EU Artificial Intelligence Act came into effect in February. This Act applies to all European Union countries. It created four risk-based categories for AI: minimal, limited, high and unacceptable.  Unacceptable AI behaviours include social scoring, the evaluation or classification of people based on their social behaviour, personal characteristics, or personality traits, with assigned scores or rankings. AI porn isn’t addressed.

EU guardrails also cover harmful use of AI biometric identification, such as fingerprints, facial features, iris or retina patterns, voiceprints, and even behavioural patterns, with the exception being for law enforcement, subject to judicial approval. Again, nowhere is there a mention of AI sexualization and porn.

Elsewhere, the U.S. passed the Take It Down Act in May 2025 that criminalizes the publication of non-consensual intimate content on the web. The act is intended to eliminate AI-generated deepfakes. It doesn’t put limits on the creation and viewing of AI pornography and erotica for personal use.

The United Kingdom has passed the Online Safety Act, which is similar in intent to U.S. efforts.

Canada has yet to pass any law, even though the Artificial Intelligence and Data Act (AIDA) was introduced to the House of Commons back in 2022, but never voted on to become the law of the land.

Australia’s New South Wales has passed legislation that makes the publication and distribution of sexualized content on the web a criminal offence. The Australian federal government has an eSafety Commissioner with the power to remove sexualized content and impose fines on its creators and distributors.

The Chinese government’s social credit system is an example of AI social scoring designed to control citizens’ ability to travel, get healthcare, education and many government services. Viewing or creating AI porn isn’t mentioned as a behaviour to be scored.

None of these countries has yet considered a sexualized Generative AI acting as a personal assistant to its subscribers. This is new territory, fraught with challenges and implications that may lead to legislation and establish new precedents

.