‘It looked real’: How AI is being used against Indian Muslim women | Islamophobia

New Delhi, India – When Samreen Ayoub saw the video for the first time, she was shocked.

The independent brand from India controlled by Kashmir was scrolling through his phone last year when a friend sent him a video of the cycle on Instagram.

Recommended Articles

list of 4 itemsend of series

The video appeared to be about his life in New Delhi, full of quotes, catchphrases and headlines as part of a TV news story. But it was a fantasy.

“It was a good follow-up,” said Ayoub, 24. “He followed my life from my first semester to my last at university.”

The video interweaves photos from Ayoub’s time as a student at Jamia Millia Islamia University in New Delhi – photos taken during everyday school activities, including group projects, farewell parties and photos taken with classmates.

The voice, which was generated using artificial intelligence, falsely claimed to be a Muslim woman “selling her body” to Hindu men. It is not clear about the people in the pictures and calling her brother who is her “pimp”.

“It looked so real that if someone, even my parents, saw the video, they would think it was real,” Ayoub said.

She is one of a number of Muslim women who have been exposed to what researchers have described as a highly visible technique: the use of AI to create sexually suggestive and misleading images.

Al Jazeera reached out to several Muslim women who were targeted. They refused to speak on the record, citing shame and the risk of retraumatisation.

‘Just thinking about sex in pictures’

The practice of sexualizing images and videos of Muslim women comes alongside India’s increasing presence in international discussions on AI governance, including the high-profile AI Impact Summit held earlier this year in New Delhi that focused on innovation and regulation.

A study by the Washington, DC-based Center for the Study of Organized Hate (CSOH) analyzed 1,326 AI-generated photos and videos collected from 297 people’s accounts on X, Facebook and Instagram from May 2023 to May 2025.

“Generative AI has made the transformation of sexual concepts into images possible quickly and at no cost,” said Zenith Khan, co-author of the study and digital research researcher at CSOH. “Graphics and depth generators allow people to turn fake news into visuals with less technology.”

Researchers are not the only ones following these developments.

Meri Trustline, an online helpline run by Mumbai-based RATI Foundation, has also seen an increase in such cases. A 2024 report by the helpline revealed this: While the media’s attention is focused on celebrities and politicians, women who are not visible in the public eye are also being scrutinized through images that, although artificially created, can cause real harm.

Frontline advisers to the aid group, including Salman Mujawar, whose work with survivors is the basis of evidence published by the agency, said they would highlight the number of such cases.

Since its inception in 2022, Meri Trustline has handled more than 482 cases, nearly 10 percent of which involved digital assets – a segment that has been growing as AI tools become more accessible.

“These violations are met with shame, fear and frustration,” Mujawar said. “Incidents are rarely disclosed even to close family members, let alone major public issues.”

‘Political crimes’

Ayoub’s video went viral on several social media accounts a few hours later. Abusive comments, threatening phone calls and accusations of his behavior quickly followed.

“It looked like a digital hack,” he said. “Not one but more than a dozen accounts posted the video anywhere, and hundreds more are sharing it.”

Many of CSOH’s creations include AI-generated memes depicting Muslim women wearing religious clothing in sexual acts and pornographic images aimed at journalists and activists. In many of these images, researchers noticed a recurring pattern: a “Muslim woman” paired with a “Hindu man”.

“In these stories, Muslim men are often portrayed as violent or corrupt,” Khan said. “Right now, Muslim women are portrayed as submissive or ‘saved’ by men from many communities.”

The images, the researchers argued, were not just about politics – they are part of it.

Sahana Udupa, a media expert at Ludwig Maximilian University in Munich, described this as part of “political prostitution” targeting women and minorities. Right-wing digital cultures, he said, combine jokes, memes and sexual images to address violence.

“The practice creates the environment,” Udupa said. “They thrive on group celebration and group violence.”

Experts say that this creation has deeper roots than misogyny. Writing in the peer-reviewed South Asian Multidisciplinary Academic Journal, researcher Soma Basu said that what is happening is the politicization of sexuality.

Muslim bodies have become a battleground for the control of all people – a force that has become more visible “Sulli Deals” and “Bulli Bai” controversial, satirical platforms targeting Muslim women in India and Basu’s association with formal support from ruling Bharatiya Janta Party (BJP) officials and informal support from party activists.

Khan’s research approaches the same point from a different angle. “In many South Asian cultures, women are shown as the honor of the family,” he said. “So attacking Muslim women appears to be a way of portraying Muslims as inferior.”

Khan also found the research to be very influential. “As a Muslim mother and a researcher, it really affected me,” she said.

“I remember being horrified when I saw pictures of a woman wearing a headscarf being represented as soft porn. As a woman, you already deal with disgusting behavior every day.

Responding to this, Atif Rasheed, a BJP politician, said AI “can be used both positively and negatively” and called for strong laws to prevent its misuse. He called the deepfakes and pornographic content “deeply disappointing” and said strict action should be taken against the culprits.

However, he refused to look at the issue through a religious lens, saying that the BJP “respects women of all religions” and the “Sulli Deals” and “Bulli Bai” cases were handled according to the law.

A common example of AI enhancement

The episodes Sulli Deals and Bulli Bai, which took place in 2021 and 2022, used popular images. All of this led to outrage and a police investigation.

Indian authorities arrested Aumkareshwar Thakur, accused of creating the “Sulli Deals” handle, and Niraj Bishnoi, who was identified as the creator of “Bulli Bai”, in January 2022. Both Thakur and Bishnoi were granted bail by a New Delhi court on “humanitarian grounds” two months later.

Researchers believe that the rise of AI has greatly increased the number and speed of harassment of Muslim women online. New apps allow users to upload photos and create sexy photos of themselves. Such tools are widely available on the Internet, often free, and require no technical expertise.

“There is a very long history of technology being used to combat violence against women, especially minority women,” said Eviane Leidig, director of research and outreach at CSOH. “What’s different is the amount of infringement and the amount of harm that AI tools can do.”

For those already living with constant harassment, the exposure of AI-generated images has added a new layer of fear.

Afreen Fatima, 27, a researcher and activist who has been harassed online since she spoke out against India’s Citizenship (Amendment) Act in 2019, was one of many Muslim women whose photos were uploaded and sold on Sulli Deals.

The law, which the United Nations has described as “discriminatory” against Muslims, seeks to expel Indian citizens from a minority of non-Muslims arriving in India from neighboring Muslim countries before 2025.

Four years after the Sulli Deals controversy, the abuse has not ended. Anonymous accounts, often using ordinary Hindu names, continue to target her with abusive messages, threats of rape and harassment, some tied to her work, although she is no longer available online.

“Every few days, there is a message from a random account threatening rape or death,” he said.

The prospect of AI-generated sexual images has added to that fear. “When I read about these pictures, I felt it was personal. It causes fear of psychosis.”

Fatima said online hate has also changed the way she visits the most visited sites.

“I don’t feel comfortable walking alone,” she said. “When you see these kinds of stereotypes circulating online about Muslim women, you start to wonder if someone would attack you in real life.”

‘I feel blue’

After the video went viral, Ayoub’s career opportunities began to fade.

He said: “As an example, your reputation is important. “If negative comments appear on your reputation, businesses will stop approaching you.”

For four or five months, fake accounts filled his profile with insulting words, putting off potential customers. The persecution also changed his relationship with social media.

“Instagram was a safe place for me,” she said. “Now I don’t feel safe there, and I put limits on what I write and how I write.”

Ayoub reported the incident to the cybercrime police in New Delhi, and filed a complaint. “Nothing happened,” he said, adding that most of the abuse was removed because his friends reported many of their accounts.

Legal experts say India’s existing laws struggle to adapt to AI-driven developments. “The damage is real even when the image is created,” said Apar Gupta, a lawyer and founder of the Internet Freedom Foundation.

Under Section 66E of India’s Information Technology Act, criminal penalties apply for capturing or publishing images of a person’s private area without permission. But if the target’s body isn’t actually labeled — in other words, if the image is AI-generated — the offer won’t work.

“Even if the image is fake, it creates a red flag for women,” Gupta said.

Digital platforms, too, enjoy “safe harbor” protection as long as they remove illegal content when notified. But Mr. Gupta said that many people affected by the accident find it difficult to get there.

“Platforms don’t make it easy for you to say this is my picture, this is serious, you have to remove it,” he said.

Without structural changes in platforms, algorithmic requirements and laws, AI-generated abuse will continue to spread faster than any laws can respond, he warned.

In such situations, accountability becomes difficult for Muslim women who are targeted.

“What I really wanted was to find people who have this account,” Ayoub said. They destroyed my reputation without even knowing me.