By Jasmin Asifiwe
Content warning: references to sexual assault and child pornography
I saw obscene child pornography on my feed today. I quickly reported it, and it was quickly deleted, but it’s not something you can erase from your mind. Technology-enabled sexual abuse is real and rising. It’s estimated that 98% of all deepfake content online is pornographic and non-consensual and that 99% of that content depicts women.
AI technologies are increasingly being used to produce non-consensual, sexually explicit images of women and children, ‘nudifying’ pictures of clothed children and manipulating photos and videos to depict women as children in sexually abusive scenarios. Common user prompts include “put her in a bikini”, “bend her over”, “make her tongue stick out”, and “remove her school uniform”.
Men on X (formerly known as Twitter) have made headlines for generating deepfake pornography with Grok – a generative AI chatbot integrated into the app by Elon Musk in 2023. In layman’s terms, billionaire Elon Musk has created a child sexual abuse generator. Grok’s ability to nudify at the drop of a hat makes it a uniquely dangerous tool.
It’s only after widespread public outcry and regulatory pressure – pressure long past the point of creating diamond – that Musk limited the bot’s ability to generate images of real people. Yet, capitalism rears its ugly head once again and the ever-erroneous Musk, culpable yet never taking responsibility, continues to allow X users paying for the subscription access to Grok’s image generation features.
The term grok was coined by American author Robert A. Henlein to mean emphasising and understanding something so profoundly and deeply, that an observer becomes part of the observed. It’s a word long known in programming and software culture for the concept of capability and understanding; naming an AI chatbot after this term carries a deep implication that X’s chatbot is a radical, truth-seeking engine. Musk’s chatbot is not ‘grokking’ women and children but reversing the term’s ethos but reducing them to datasets to be transformed at will – even if it means violating their dignity. The human being behind the image disappears.
AI is by no means a neutral technology: it scales sexual exploitation. It shouldn’t fall completely on these artificial entities – the men using artificial intelligence in this way are real people with real victims. Ideological signalling aside, the imposing AI monopoly plagues us, and over half of Australian adults fear AI being used to cause them harm.
It’s not a phenomenon. This non-consensual sexualisation is representative of] the greater institutional systems at play, where Ghislaine Maxwell is still imprisoned for her role in Jeffrey Epstein’s rape and pillage of children, but Donald Trump and the rest of the global elite continue to hold high-yielding positions and power. The message is reinforced that accountability is disproportionately applied. One woman becomes the face of punishment, creating the appearance of justice having been served, while the wider constellation of male influence remains untouched.
In this sense, the persistence of technologies that enable sexual deepfakes, are digital extensions of a broader network of complicity.
The question then is how best to confront the structures that allow exploitation to persist. A South Australian teenager has become the first person to be prosecuted under new nation-leading federal laws, aimed at combating deepfake pornography. Strict age-verification requirements now block access to pornographic websites. But maybe something more intrusive – radical even – is vital, as the weight of the crime is everlasting. You can’t redress those who’ve already been exposed, exploited, and put on display.
