Pakistan’s AI trust challenge: Can it be fixed?

ISLAMABAD, (APP - UrduPoint / Pakistan Point News - 29th Jun, 2026) Artificial Intelligence is rapidly transforming life in Pakistan, making work, education and communication more accessible. But the same technology is also eroding one of society's most valuable assets: trust. AI-generated deep-fakes, manipulated voices and fabricated images are making it increasingly difficult to distinguish fact from fiction. In a country with around 110 million internet users, the misuse of AI is growing much faster than the laws and institutions meant to contain it.

The growing crisis is perhaps most visible in the targeting of women.

For Nighat Dad, founder of the Digital Rights Foundation, the rise in AI-generated harassment is exposing deeper societal problems. The organization reported 3,171 new cases involving AI-enabled manipulation of images and voices in 2024. Yet Dad believes these figures only scratch the surface.

"Pakistani society is patriarchal, it won't believe the woman," she says, explaining why many victims never report such abuse. In her view, AI has simply handed harassers more convincing tools, while the burden of proving innocence still falls almost entirely on women whose images and voices have been manipulated.

The challenge extends beyond removing fake content. It is about restoring credibility once trust has already been damaged.

That concern is echoed by Sadaf Khan of Media Matters for Democracy, who says the consequences of AI-generated deep-fakes go far beyond online embarrassment.

"Once a woman is targeted this way, her image is seen as immoral, and the honor of the entire family is lost," Khan says. In Pakistan's patriarchal honor culture, she argues, even after manipulated content is exposed as fake, the social damage can linger for years, putting victims' safety and livelihoods at risk.

The vulnerability is not limited to ordinary citizens.

Punjab Information Minister Azma Bukhari also became the target of a deep-fake after her face was used in a fabricated online video. "I was shattered when it came into my knowledge," she later said. She continues to litigate the matter in the Lahore High Court. Her experience illustrates that if a senior member of the provincial government can become a victim of AI manipulation, an ordinary Pakistani with fewer resources and less public visibility is likely to face even greater challenges in seeking justice.

Official figures reveal how difficult accountability remains. The FIA received roughly 1,180 complaints involving deep-fakes and non-consensual imagery in 2023, most of them from women.

<?php /*?> <?php */?>

Yet only 12 cases resulted in charges and just seven ended in convictions.

The trust crisis extends beyond adult victims. Pakistan is also part of a rapidly worsening global trend involving AI-generated child sexual abuse material. A 2025 study by the Childlight Global Child Safety Institute found Pakistan hosted more than one million reported cases of such material in 2024, among the highest volumes in South Asia. The study also found AI-generated abuse imagery worldwide surged by more than 1,300 percent in a single year, driven by tools capable of creating convincing explicit images of real children using only a handful of ordinary photographs.

UNICEF's 2026 study across 11 countries found that roughly one in 25 children had already had their photographs manipulated into deep-fakes.

Commentators in Pakistan have linked this emerging threat to the tragedies that originally shaped the country's child protection laws, including the 2015 exposure of a child abuse network in Kasur and the 2018 murder of seven-year-old Zainab Ansari, both of which helped drive the passage of PECA. Legal experts are now calling for the law to be amended again so it explicitly criminalizes AI-generated and AI-manipulated child abuse content, arguing that existing legislation was written for authentic photographs and videos rather than synthetic media.

Teenage girls and young women face another growing danger: AI-assisted image-based extortion. Criminals increasingly use real or fabricated explicit images to blackmail victims. FIA cybercrime records document a steady stream of such cases, including an 11-year-old girl in Punjab blackmailed for months through a fake TikTok account, a minor in Faisalabad threatened with having her photographs sent to her parents, and a Peshawar man arrested after allegedly hacking a minor's Snapchat account to extort her.

Artificial intelligence is not the real threat. The greater danger is a society where people can no longer trust what they see, hear or share. Every convincing deep-fake doesn't just deceive one victim, it chips away at public trust, making it harder to believe genuine evidence, real victims and even democratic discourse.

Preventing Pakistan's AI trust crisis will require more than better technology. It demands stronger laws that recognize synthetic media, faster action by law enforcement and online platforms, and widespread digital literacy so people learn to verify before they amplify. Because in the age of AI, rebuilding trust may prove far harder than losing it.