Reena Wazir, a student at the University of Punjab, was scrolling through social media one night when she came across a video which shocked her. At first, she thought it was real. The clip appeared to show scenes from a drama, but something about it felt wrong. She watched it again, only to realise that it was manipulated. Soon after, a discomforting thought started lingering in her mind that what would happen if this video was about her..
“I don’t understand what it [Gen AI] is capable of doing. When I see what happens to celebrities, [their deepfakes are circulating online] I think, if they can’t control it, how can I?” she remarked.. “These are actresses. Their families are educated. People know such things can happen to them. But what about us?”
Wazir hails from Waziristan and is actively involved in student politics. For her, the fear was not simply about having a fake video made using her image. It was about what would happen after it was created, especially in a society which is quick to judge a woman’s morality without asking if something is true or not. “If something like this is made using my picture, how will I explain it?” she said. “Even if my family understands, the wider community will not.”
A leaked image or manipulated videos do not only affect the woman in it. In Pakistani context, it lands on her entire family, and sometimes community, notes a June 2026 research report by Chayn, a global tech non-profit empowering women and marginalised genders facing abuse. The report noted fathers of survivors (of image-based abuse) were unable to work, marriage prospects of siblings were impacted, and households were looked upon in a “shameful” manner.
The concerns expressed by Wazir also come from lived experiences of people from her region. For example, the killing of University of Lahore student Sunaid Dawar, also the chairman of the Pashtun Council, in 2022 has made her fearful. Dawar was a staunch advocate of girls education in Waziristan.
Asma Gull, another student leader from Waziristan, has also limited her social media presence because of similar concerns. “I haven’t completely stopped posting,” she explained. “It’s necessary for politics. But sometimes after uploading something, I can’t stop thinking about it. I’ll even delete it later.” Then she paused and added “I feel like I am back where I was five years ago. Like I am again in my hometown. But this is more terrifying, because that fear was about the present. This one is about the future too.”Wazir pointed to the political uses of such technologies. “In the past, militants used edited photos to label people as immoral, enemies of Islam or foreign agents. They were successful. If they start using generative AI, the consequences will be much worse”.
TFGBV and its impact on women
“Generative AI is accelerating and expanding the ways in which technology-facilitated gender-based violence is being used to target women and gender minorities, especially as AI generated content is becoming increasingly realistic,” shared Shmyla Khan, an Aurat March Lahore organiser and former Director of Research and Policy at the Digital Rights Foundation.
In a place like Pakistan, where the easiest way to discredit women is by attacking her reputation, such images can often be very effective in targeting women engaged in political work.
The emergence of generative AI, combined with issues such as image- and identity-based abuse, is unfolding within societies like Pakistan where women’s visibility is already contested. Women’s political participation remains limited, and societal anxieties around women’s sexuality continue to shape public perception often with violent consequences. In such a context, the expansion of technological capability does not necessarily translate into empowerment. Instead, it produces new forms of caution, withdrawal, and silence. This has direct implications for women in student politics.
According to the Digital Rights Foundation’s Annual Report 2025, the organisation received 361 complaints related to image-based abuse, indicating that such harms are increasingly becoming part of lived realities in Pakistan. Reports of identity-based abuse where publicly available information is used to target individuals have also been recorded separately, pointing to a broader ecosystem of digital harm.
Double-edged sword
The political journeys of both Wazir and Gull are shaped by displacement, violence, and survival. During military operations in Waziristan they lost many close ones, and moved to Lahore along with some of their family members. During this, they encountered the Progressive Students Collective (PSC), an organisation whose politics resonated with their aspirations around gender, class and ethnic equality. Over time, they carved out space for themselves within the organisation’s leadership and gained recognition in broader progressive circles in Lahore.
“Women in politics are unimaginable in our society,” Gull said while speaking specifically about Waziristan. The general society is quite conservative, but not everyone. The issue is with the hate and militant groups which enforce conservatism under the barrel of a gun if they see any ‘deviance’, she explained. “That’s why we couldn’t even imagine doing politics there.”
Social media plays a central role in student politics; it helps with mobilisation, and becomes a form of currency for activists individually,
Moving to Lahore, however, created new possibilities. “We felt a responsibility to speak on behalf of those of us who cannot. My immediate family, particularly my brothers, supported my political engagement. But the wider social environment remained restrictive. We still had to self-censor in many ways,” Gull shared. This self-censorship is not only external; it has been internalised over time. “There is this muscle memory of being silent,” she said, adding that she did overcome these fears eventually. “I started speaking publicly, participating in discussions, and eventually sharing my political activities online too.”
Social media plays a central role in student politics; it helps with mobilisation, and becomes a form of currency for activists individually, Gull said. Online visibility functions as an informal certificate of activism. In many cases, it carries more weight than the actual labour of organising. People like us who are actual political workers cannot leave this space. “But the emergence of generative AI has halted that progress because it brings so many threats,” she added.
Wazir, on the other hand, shared an experience in which she experienced how easily meaning behind a social media post can be distorted even in her social setting. “Once I posted a picture holding a microphone at a political event of my organisation in Lahore. My relatives showed it to everyone in my hometown and family, saying I was singing at a concert in Lahore and used extremely inappropriate words for me. My family was very disturbed by this. Now imagine if it was an AI-generated fabricated video. What would my relatives make of me instead of a singer?”
At the heart of these fears is a crisis of credibility. In environments where technological literacy is limited, the distinction between real and fake becomes difficult to sustain. “If my father or relatives saw such a video, they wouldn’t think it’s AI,” Wazir said. “They would think it’s real. How do you explain something like this?” Even when explanations are possible, they are not always accepted. Authority structures within families and communities do not necessarily accommodate such explanations.
“They don’t listen to younger people,” she explained. “They will say, ‘If it’s your picture, how can someone else create it? My brothers are educated and would definitely believe me not just because they are progressive but also because they are aware of these trends. But my other family members never would.”
In such contexts, the burden of proof falls disproportionately on women, who are already subject to moral scrutiny. And even when they are not at fault, they are often blamed. “Some would accept it’s not real,” Wazir said, “but they will still blame me as society can find excuses immediately to blame women. In this case it can be like why the other person who created the content had my picture in the first place? Why did I upload it online? Why did I even have social media accounts?”
Lack of media literacy
Wazir pointed to the need for immediate and collective responses to counter the harms of generative AI particularly in the context of image-based abuse. “There is no public awareness about this. People cannot even identify AI-generated content. We need discussions, workshops, seminars. Government and non-governmental organisations working on gender and technology should play their role in this regard,” she said.
She also emphasised that political organisations need to rethink how they evaluate participation and overall contributions of their members. “If visibility becomes risky, then we cannot measure commitment only through online presence,” she said. “There has to be some adjustment like there should be a count of physical participations, length of membership and designations should be given on the basis of that original political work.”
Perhaps the most immediate step, however, is recognising the legitimacy of this fear.
“Yes, of course there is fear,” Gull remarked. “Even if nothing has happened, it is always there.” That fear, persistent and anticipatory, is already reshaping behavior. It determines what is shared, what is deleted, and what is never posted at all.
“The solution to these issues is for TFGBV to be addressed through robust and survivor centric reporting mechanisms that are responsive to the needs of victims. These systems need to be agile enough to recognise the unique challenges that Generative AI poses in exacerbating the problem,” Khan said,
Chayn’s indepth report on the matter also highlights that a major part of the problem lies in the policies of social media platforms, whose Western-centric definitions of intimacy and lack of contextual understanding of local languages slow the process of content removal. Khan also highlighted a broader need to recognise the work of women in politics by challenging notions of respectability politics that allow women’s lives to be disrupted based on both fake and real images being weaponised against them. Alongside the need for cultural and political shifts,
Understanding Image-Based Abuse
Image-Based Abuse is a sub-category of Technology-Facilitated Gender-Based Violence (TFGBV) which covers a range of harms carried out through digital tools. The discussions surrounding image-based abuse are often on explicitly sexual content and it’s often used interchangeably with phrases such as non-consensual intimate imagery (NCII), and intimate image-abuse.
Image-based abuse is commonly understood as the creation or threat of sharing intimate or sexual images, whether real or deepfake, without a person’s consent. However, a recent report by Chayn challenges this narrow definition, arguing that the phenomenon extends beyond “non-consensual nudes,” because notions of nudity and intimacy vary across cultural contexts. Even seemingly ordinary images can be manipulated or recirculated in ways that render them harmful without being explicitly sexual.
