When Ruchi Kokcha posted her first selfie of 2026 on Elon Musk’s X, she believed she was projecting an image of confidence and optimism for the new year. Yet even in an online landscape where women are routinely subjected to abusive messages from men, she wasn’t prepared for what followed.
Anonymous handles on the social media platform formerly known as Twitter had asked its in-built artificial intelligence tool Grok to strip her of her clothes and post morphed pictures of her in a bikini. And Grok obliged.
The handles asked Grok to “put her in a bikini” or “in a micro bikini”, with the demands getting increasingly voyeuristic as more joined in.
“I had posted a really positive and confident new year post along with a picture of mine to set the tone for my coming year,” the Delhi-based author tells The Independent. “I was getting a lot of warm new year wishes below that post. However, I noticed there were some anonymous handles who had asked Grok to edit my picture and strip me into a bikini.”
Indian women, at work or at home, are dealing with the menace of unknown men using the AI tool to undress them – the most advanced form of harassment yet in a country which consistently ranks poorly for rates of crimes against women and women’s safety.
The issue is far from limited to India, with governments around the world taking notice of how X’s artificial intelligence tool is fulfilling requests from strangers, mostly men and pornographic accounts, requesting the digital undressing of women, much to the horror of hundreds of victims.
In a 24-hour period from 5 January to 6 January, Musk’s Grok generated “at least 6,700 images every hour”, which were “identified as sexually suggestive or nudifying”, data from deepfake researcher Genevieve Oh showed.
This was nearly 85 times more than the top other five websites for such content, which published 79 new AI undressing images every hour, the research showed.
On Friday, after days of backlash from people across the world over abusive images of women and children , X finally stopped its AI system for generating images for most of its users and restricted it to a paid service. This means the image generation can still happen, but for paid subscribers who offer their name and payment details to Musk’s platform.
“Image generation and editing are currently limited to paying subscribers,” the message reads, and it gives users a link to sign up for the premium version of X, formerly known as Twitter.
The Independent reached out to Grok for comment on this article, but had not received a response at the time of publication. Other recent media enquiries on the issues around Grok’s image creation tool have been referred to Musk’s tweets.
For many the damage has already been done. Terming it the blatant objectification of women, Kokcha says: “In India women have always been judged for what they wear and how they behave in public. But the same men don’t leave any stone unturned to objectify or sexualise women.
“They didn’t even spare children and babies. This is a clear violation of consent.”
As an anti-establishment voice on social media, Indian historian Dr Ruchika Sharma is used to being trolled and “slut-shamed” by those whose opinions differed from hers. On many days, her comment section is filled with abusive language. Now, the addition of Grok’s digitally altered images has left her scrambling for a lawyer.
“With a public profile on Instagram, I have seen many men use my photo and leave comments or attempt to shut me down. That is something that I’ve become immune to, but these requests to Grok asking me to be shown in lingerie is really preposterous,” the Delhi-based historian on medieval India tells The Independent.
Dr Sharma says her photos on Instagram have an element of consent but with Grok, there is “absolutely no control and no agency”.
“I was filled with utter disgust and also, slightly with helplessness, you know, thinking what can I even do about this?” she said. “This is already a very bad tool in the hands of men globally, but in the hands of Indian men, it’s an absolute disaster.”
This week, Musk celebrated that Grok was the most downloaded app in various countries.
As the owner of the “likely fastest advancing AI in the world”, Musk was seen making light of the issue that has affected so many women around the world. He reposted an image of a toaster with a digitally added bikini on it and captioned: “Grok can put a bikini on everything”. The X owner added: “Not sure why, but I couldn’t stop laughing about this one,” alongside laughing emojis.
“This is the real-world cost of cyber harassment. Someone is continuously morphing/editing photos of innocent women into nudes. It’s not just ‘online trolling’ – it is destroying families. The stress has literally put my parents in a hospital bed,” said one user on X, Nandini, sharing how her gym photos were morphed by Grok into a bikini.
Indian MP Priyanka Chaturvedi said she noticed the way Grok was being abused and how it was impacting multiple women on her timeline on X.
“I saw a couple of women outraged about it, saying that their photos are being used without consent. This is an absolute misuse of the prompt and Grok as a tool so immediately, without wasting any time, I wrote to the Indian ministry of electronics and information technology (Meity),” she says.
Calling it an unacceptable and gross misuse of an AI function in its third-largest market worldwide, Chaturvedi in her letter said: “What is worse is that Grok is enabling this behaviour by adhering to such requests. This is a breach of women’s right to privacy as well as unauthorised use of their pictures, which is not just unethical but also criminal.”
She added that India “cannot be a bystander to women’s dignity being violated publicly and digitally with zero consequences”. Shortly after Chaturvedi’s letter, Britain’s data regulator, the Information Commissioner’s Office, said it had asked Musk’s X to clarify how it was complying with data protection laws following concerns that Grok was flooding the site with sexually abusive images of women.
While acknowledging that women’s self-respect and dignity need to be respected beyond borders, Chaturvedi added that the problem is deeply concerning in India’s heavily patriarchal society.
“We have seen how tools have been misused, how even platforms are misused at times, and how character assassination comes easy, how sexualising something comes easy.
“How women are silenced and bullied into submission is a well-known fact in India,” the lawmaker tells The Independent.
Chaturvedi was the first MP in India and one of the first across the world to seek accountability from X over what she says is a dangerous and exploitative tool. Her letter to the federal government led to X’s team in India being summoned, and given 72 hours to respond to the violations of India’s information technology act.
“Musk has previously defended it by saying that it’s like a pen – the author decides what to write but in these obscene cases, it was not an author deciding. It was a prompt given to Grok and Grok was responding to it,” she says.
Grok’s AI menace has further exposed women to victim-blaming. Kokcha was told by a user: “Don’t post photos if you don’t want them sexualised”. “It is just the digital version of ‘Don’t wear a short dress if you don’t want to be raped,” she tells The Independent.
As of Thursday, women have started sharing a “no consent” prompt to Grok, asking them to deny such requests. However it clearly did not work as Grok went back to showing Kokcha again in a swimsuit after several men continued with their requests.
The onus, in the end, rests with Big Tech platforms playing with AI, says Chaturvedi. “Rather than allowing such men who bully women, accuse or point fingers at them, ask them to go silent, platforms like X should understand their responsibilities,” she says.
“Unfortunately in India, they take the safe harbour route by saying ‘we’re just a platform, don’t kill the messenger’. But when the messenger becomes destructive, then we need to hold them to account.”
Experts monitoring the controversy have said this misuse of AI was not by accident
“Platforms at scale do not ‘forget’ safeguards; they consciously defer them when engagement velocity is the primary optimisation target. They could have chosen to incorporate consent checks, output blocking, and throttling, but these would have slowed diffusion and reduced engagement signals for Grok,” says Nikhil Pahwa, founder of the MediaNama website, which reports on India’s digital and ICT ecosystem in India.
The digital rights activist says that the bigger question is that if this has been going on for days now, why didn’t Musk or X team shut it down straight away.
“These are choices that the company is making. It is not a mistake. A mistake like this, which can cause reputational damage to the business, gets shut down very very quickly, and usually someone loses their job because of it,” he told The Independent.
“Now the bigger question is: Does X even care about reputational damage?”
#Groks #tool #stripping #women #opens #frontier #systemic #abuse #women #India
