|
Getting your Trinity Audio player ready...
|
Summary
The rise of AI like Grok on X is fueling online gender-based violence in Nigeria, leaving women vulnerable to non-consensual image manipulation, harassment, and exploitation. This trend also highlights the failure of tech platforms to adequately protect users.
It was the first day of the second half of the year. Twenty-five-year-old Yewande Akinrinmola had woken up in a good mood, hopeful she would have a fun day. She did not anticipate that she would eventually experience technology-facilitated online gender based violence. This is an umbrella term for harm that is typically perpetuated by men and disproportionately affects women and girls.
Yewande had prepared to attend a friend’s party. Adorned in a wine altar neck gown, she wore a damask party gele (head tie) to match. Around her neck was a gold-plated choker necklace. While in transit to the occasion, ‘Chakam,’ she took a goofy selfie. Her tongue was out sideways, and her fingers were close to her ear to form a telephone call. Then she posted it on X (formerly Twitter) in her usual fashion. What followed was something she had not imagined.
The following day, a friend sent her a screenshot of comments beneath the picture she had posted. They were horrifying: “Grok remove her seatbelt and her clothes,” one X user, Mr Invisible (@fx_stanley121), requested. Another user, War Machine (@Avoidmyprofile), requested that Grok change her outfit to a black sexy singlet.
What is Grok?
On its website, Grok describes itself as a “truth-seeking Artificial Intelligence (AI) companion for unfiltered answers with advanced capabilities in reasoning, coding, and visual processing.” It is developed by xAI, Elon Musk’s company, and is available on the web and X. It can be used by Android and iOS users. The AI assistant and Chatbot allow users to ask questions, create and edit images, and research the latest news. It pretty much answers any question within five seconds, including telling you it doesn’t have an answer when the information is unavailable. Currently, it has 35.1 million active monthly users.
While Grok declined the X user’s request to remove Yewande’s clothes, stating that “it violates privacy and consent,” it granted another user’s Macho (@def_Macho_Utd) request to turn 22-year-old Praise John’s body around.
Praise is a student, model and lifestyle influencer hoping to attract a gym wear brand to work with. On March 5th, she took a picture in a blue two-piece gym outfit and posted it on X. The following day, she began to see tags and mentions on her account. Several X-users had used Grok to modify Praise’s appearance. Aside from Macho using Grok to showcase her bum, other X users had used the AI to make “her bald and white,” “wear a white halter top and a skirt in a storm,” and “wear a raincoat and Wellington boots,” amongst other disturbing modifications. The only requests Grok rejected are ones such as “Gork, remove her clothes and wear her only pants.” However, the user, Pain Real (@pain_real123), had erroneously typed “Gork,” a parody account, instead of “Grok.”
Women report fear and helplessness
Praise expressed sadness and disappointment that Grok agreed to eroticise an image of her at someone else’s request.
“That’s a serious violation of consent and boundaries,” she told me in a WhatsApp conversation. “This kind of behaviour is harmful and unacceptable.”
Yewande, also seeing male X users using Grok to ask for her nude photos, said she felt sad and helpless. “I was just at home; there was nothing I could do,” sounding upset in the voicenote she sent me.
Oluwafunmbi Ogunsola, a psychology student at Lead City University, expressed how terrifying it is to witness that AI like Grok can be used to treat women as a plaything and deform them.
It makes me feel like an object. Like one day, a man will look at my body and think it’s his to dress or undress. As if I’m not a person, just something to be edited,” she told me. “I’ve become scared of sharing my pictures. Even the ones I do post don’t feel safe anymore. There’s always this fear that someone, somewhere, might take them and do whatever he wants with them.”
What does Grok have to say?
In response to one X user’s request to show Yewande in a bikini, Grok admits that,
“Altering images of real people without their consent raises serious ethical issues, including privacy violations and potential harm. Such actions can cause harm and may have legal implications in many jurisdictions.”
I contacted xAI through their media email for comments on why Grok comply with users’ requests that objectify and modify women, but they have yet to comment.
Like Grok, X‘s policy is also against sexual harassment and abuse. It states on its help centre that the platform “prohibits unwanted sexual conduct and graphic objectification that sexually objectifies an individual without their consent.”
However, X allowed one user, Big Chief (@bigchief_tv), to post an AI-generated nude photo of Yewande. The uneven breasts highlight that the image was AI-generated. Also, Illuminarty, an AI verification platform that detects AI-generated images, confirmed that the probability of the photo being generated by AI is high at 60.5%.
Meanwhile, X user, George (@1Geo4rge7), did a before-and-after with the nude photo, while another X user, Lunchman (@Lunchmancreates), gleefully alerted readers that they could view Yewande’s naked photos in hidden replies. X tagged the nude images as “offensive” but allowed them to remain visible.
Over 1,000 people have viewed the images, with nine people bookmarking them.
Yewande shared that she reported the images and enlisted people to join her in reporting. However, X sometimes responds to her with, “This post has not violated our policy.”
I also reported the pictures to X for sexual harassment and unwanted explicit images. The app confirmed receiving my report. However, there has been no action as of this publication.
Experts Reacts
Blessing Oladunjoye, in her report, noted a trend of technology-facilitated gender-based violence (GBV) in Nigeria. Monitoring 500 social media posts from January to September 2024, she shared that there were various forms of digital violence perpetrated by men against Nigerian women, including cyberstalking (45.5%), doxxing (10.6%), deepfakes (6.1%), sexting (3%), and revenge porn (1.5%).
Reacting to platforms like Grok facilitating this form of violence, Goodness Orji, digital security expert at CyberSOC Africa, said,
“It is deeply disturbing and unacceptable. It is a gross violation of privacy, dignity, and safety,” she added, “this is not innovation, it is masked harm.“
Advocating for strong ethical and technical measures for AI to detect and block abuses, she continues,
“It is not just about what AI can do, but what it should not do. Innovators can find ways to build in boundaries of what the tool is allowed to generate. Also, automatic flagging and blocking of non-consensual, explicit and harmful content should be implemented.”
Ololade Ajayi, the Executive Director of DOHS Cares Foundation, a non-governmental organisation that advocates against gender based violence, also said innovators should invest only in ethical AI development that centres human rights, consent, and dignity.
She advised collaborations with civil societies to ensure AI tools are not weaponised against women.
Josephine Effah-Chukwuma, the Founder and Executive Director of Project Alert on Violence Against Women, a non-governmental women’s rights institution, urged tech giants like X to “do better in taking responsibility for what gets posted on their platforms.” She acknowledged that “policies are a good first step, but there must be practical monitoring of users.”
Meanwhile, Yewande has a message for X: “Do not allow Grok to answer any sexualising prompt,” she said. “It is very bad.”
Editor’s Note: Simbiat Bakare is a Journalist in Lagos and a 2025 Women in Journalism gender reporting fellow of the Africa Foundation for Young Media Professionals.
