Sexual offences lawyer warns of 'explosion' of illegal images as ChatGPT plans adult content
As artificial intelligence has become more powerful and prevalent in our everyday lives, it was perhaps only a matter of time before it was used to create adult content.
The risks associated with using AI to generate images were made abundantly clear earlier this year when Elon Musk's Grok programme published thousands of inappropriate pictures, undressing women, covering people's faces with 'glue' and even publishing sexualised images of children.
Despite this, OpenAI CEO Sam Altman has regularly reinforced his plans to introduce erotic content on his company's popular programme, ChatGPT.
Why is ChatGPT introducing adult content?
In short: money.
Online pornography has become one of the biggest businesses in the world, with the popularity of sites such as OnlyFans emphasising just how much cash some people are willing to part with, both for access to explicit pictures as well as mere companionship amid a loneliness epidemic.
Advert
Speaking in October 2025, Altman first confirmed ChatGPT's plans for erotic content.
"In December, as we roll out age-gating more fully and as part of our 'treat adult users like adults' principle, we will allow even more, like erotica for verified adults," he said at the time.
However, earlier this month, it was confirmed that the plans had been put on hold, as the tech company with over 900 million users claimed to be focusing its efforts on 'gains in intelligence, personality improvements, personalisation, and making the experience more proactive'.
Altman added, though, that he still believes in the 'principle of treating adults like adults', just that it will take a bit more time to get there.
Major questions now remain following the raft of new laws introduced under Online Safety Act 2025 as to what it means for the UK, which already announced bans on certain types of graphic, violent porn.
And earlier this year, X was forced to act in a bid to prevent those under the age of 18 viewing certain content as well as the tech being used to generate illegal images.
Altman is clearly keeping the door open for AI-generated adult content in the future, with it already readily available in other dark corners of the internet, but a sexual offences lawyer has now told LADbible how technological advancements in AI has the potential to cause an 'explosion' of illegal images in the future.
How could AI's adult content lead to more illegal images?
Marcus Johnstone, Managing Director of PCD Solicitors, specialises in criminal defence for those accused of sexual offences, and warns 'we must not underestimate the dangers of AI and AI-image generation'.
"Over the past two years alone, I have seen dozens of cases involving AI-generated indecent and illegal images of children, including cases where teenagers have naively but illegally generated illegal images of friends, classmates and teachers," he tells us.
"What we are about to see over the next few years is an explosion in AI-powered illegal images offences - including deepfake sexual images of adults, children and child sexual abuse. In my view, this will soon become one of the most common sex crimes, with thousands-upon-thousands of cases a year."
And while AI platforms could face the same punishments as online porn sites who do not adhere to the new age restrictions rules, the encrypted nature of messaging sites mean that it is very tricky for police to track down offenders without going undercover.
Johnstone explains: "Over the past 15 years, social media has entirely transformed the way sexual offences are conducted. The growth in social media platforms and the explosion in porn sites has upended this entirely, and online sexual offences now vastly outpace the number of sexual offences taking place on the streets.
"Recent research indicates that 90 per cent of all sexual offences involve online activity, the most common being online child sexual exploitation, grooming and indecent images of children.
"Social media, AI and the development of Chat GPT now represent another game-changing cocktail. They provide the perfect environment for dangerous predators to target children, but also mean a great many people are at risk of committing crimes themselves. I have dealt with many, many cases where naive and vulnerable people - including teenagers and children - have fallen afoul of the law without any intention of doing so.
"This includes instances where people have viewed illegal content hosted on legal porn sites, have exchanged explicit photographs and messages with children their age, or have inadvertently and unwillingly received indecent images from others. Each of these scenarios meets the threshold for a serious sexual offence, and AI will take this phenomena and grow it hundredfold."
What has OpenAI said about the concerns?
Speaking to LADbible, OpenAI made it clear that 'age prediction' has already been rolled out across its platforms, which ensures that anyone under the age of 18 will 'automatically receive additional safeguards to reduce exposure to sensitive or potentially harmful content'.
In cases where users don't disclose, or potentially lie about how old they are, age prediction works by monitoring 'behavioural and account-level signals', including how long the account has existed, the time of day it is most regularly used and other patterns over time.
Although adult mode will still be shipped, the company wants to ensure that it completely reflects the principles that it has set out and works in exactly the way it is intended.
That also means that its usage policies are designed in a way which prohibits outputs to sexualise children in any way, while training datasets are screened to detect and remove any illegal material.
An OpenAI spokesperson said: “We are committed to preventing any attempt to use our models to support child sexual exploitation and abuse, and models are designed to refuse this content. We actively detect, block and report attempts to generate or upload CSAM, including reporting content to child protection organisations and banning associated accounts.”
The tech company has promised to train models responsibly and regularly monitor any attempts to generate sexualised content involving minors, with any identified users to be reported to the National Center for Missing & Exploited Children (NCMEC).
Topics: AI, Artificial Intelligence, ChatGPT, Technology
James Moorhouse