Article Highlights
- Pakistani youth are earning thousands of dollars by creating AI fake videos targeting US and UK audiences.
- Facebook’s monetized accounts globally jumped from 2.5 million to 14 million in just one year, a rise of over 500 percent.
- More than 10,000 Pakistani accounts are active in Facebook’s content monetization program, many of which publish in English.
- During the 2025 Pakistan floods, AI-generated videos exploiting women in affected villages were widely shared online.
- Experts warn that once platforms develop their own AI video tools, millions relying on this income could pivot to fraud and scams.
When I first started looking into how young people in Pakistan were making money online, I expected to find the usual stories about freelancing or e-commerce. What I found instead was something far more unsettling. A growing number of people, many of them without university degrees or English fluency, had quietly built an income stream out of creating AI fake videos in Pakistan, videos designed not to inform but to provoke, mislead, and earn.
This is not a fringe activity. It has become, as researchers and digital rights experts now openly call it, an emerging industry.
What Is AI Slop and Why Does It Spread So Fast
The term “slop” refers to low-quality, mass-produced content generated by artificial intelligence. It is not designed to be accurate or original. It is designed to get attention. Think of it as the digital equivalent of junk food: cheap to make, engineered to be addictive, and potentially harmful in large quantities.
Creating this kind of content requires nothing more than a smartphone, a free AI tool, and a rough idea of what gets people angry or emotional online. Videos can be made in minutes. A face can be swapped. A voice can be cloned. A political figure can be made to say things they never said.
The reason AI fake videos in Pakistan and similar countries are spreading so rapidly comes down to two things: a very large young population hungry for income, and truth social media platforms that pay more when content gets more attention, regardless of whether that content is true.
| 500% | 10,000+ | 55% | $4,000+ |
| Rise in monetized Facebook accounts globally in one year | Pakistani accounts in Facebook’s monetization program publishing Urdu content | Pakistani Facebook page admins publishing content in English | $4,000+ Potential earnings from a single viral video with US-based views |
The Money Behind the Misinformation
I spoke with a Pakistani content creator named Muhammad Abdullah, who had built a livelihood from AI-generated videos. He had not completed higher education and relied on YouTube and Facebook as his main sources of income. He learned everything either from free YouTube tutorials or by paying for courses sold by others in Pakistan, often with screenshots of thousands of dollars in earnings as advertisements.
The economics are straightforward. Social media platforms pay based on views, and views from the United States generate far more money than views from Pakistan or India. A one-minute video with four million views, mostly from American users, can earn between $4,000 and $5,000. That kind of money is transformative for a young person with few other options.
“We are not interested in news. I do not even see what is being said in the videos, what is written and what is not written.”Muhammad Abdullah, AI content creator, Pakistan
He was not the only one. The Bureau of Investigative Journalism documented a Pakistani citizen who earned thousands of dollars by pushing racist AI videos to British audiences. One video falsely featured British Prime Minister Keir Starmer and sparked real anger about immigration. The man admitted in an interview that he had no idea what was going viral or why. He was simply following the algorithm.
Abdullah told me that one person in Pakistan could be running ten to twenty channels and accounts simultaneously. “They find easy and quick ways to make money with the help of AI,” he said. “They do not care about quality, nor do they care whether the content is showing morally objectionable things.”
Why Pakistan Is Particularly Vulnerable
Victor Rio, executive director of What to Fix, a nonprofit that researches technology company policies, put it plainly: social media platforms built this problem. Their monetization systems reward content that triggers strong emotional reactions, which naturally pushes creators toward content that is shocking, controversial, or misleading. AI makes producing that content faster and cheaper than ever before.
When someone in a Pakistani city sees a neighbor claiming to earn thousands of dollars from Facebook, the barrier to entry feels low. All you need is a phone. The result is that the trend of creating AI fake videos in Pakistan has accelerated faster than any regulatory or educational response could match.
According to data shared with the BBC, in January 2025, there were about 2.5 million monetized accounts globally on Facebook. By the time research was conducted for this piece, that number had risen to 14 million, an increase of more than 500 percent in a single year. Pakistan contributes a significant portion of this growth, with over 10,000 accounts publishing Urdu content and many more publishing in English, Spanish, Portuguese, Hindi, and Arabic, often using AI to translate and localize content without ever understanding what it says.
The Human Cost: Women, Floods, and Exploitation
The consequences of this trend are not abstract. During the devastating floods of 2025 in Pakistan, the Digital Rights Foundation identified a disturbing pattern: AI-generated videos showing women from flood-affected villages were being created and circulated widely online. These women had not consented to appear in any video. Their faces and situations were being exploited for engagement and monetization.
Nighat Dad, executive director of the Digital Rights Foundation and a member of Meta’s oversight board, described AI slop as an industry in Pakistan. She said the speed and accessibility of AI tools have turned it into a digital economy that many young people now depend on, even if they do not fully understand the harm it causes.
“These laws are used for purposes and purposes. This trend would not have grown so rapidly if content creators knew that creating AI slop could get them caught.”Nighat Dad, Digital Rights Foundation
She has repeatedly raised these issues with Meta in oversight board meetings. Her concern is not just about content moderation. It is about governance, accountability, and the fact that vulnerable people, particularly women and children, are bearing the cost of a system designed to maximize engagement.
Deepfakes of Real People and the Political Danger
Some of the most alarming cases of AI fake videos in Pakistan involve real political figures. A deepfake video circulated on TikTok showed Donald Trump claiming that Imran Khan was a dangerous leader. Another older deepfake had Trump promising to help free Imran Khan after a supposed election victory. Both videos had millions of views, likes, and shares, a strong indication that many viewers took them as genuine.
Israeli Prime Minister Benjamin Netanyahu publicly claimed that disinformation campaigns using “bot farms” and “fake addresses” could be traced back to Pakistan, giving the example of text messages claiming to be from Texas residents who had turned against Israel. Whether or not these specific claims are accurate, they reflect a growing international perception of Pakistan as a source of organized AI-driven disinformation.
Syed Najeebul Hassan, a representative of Pakistan’s National Crime Agency (NCIA), acknowledged that the problem has become serious. He said institutions are fighting on two fronts: organized criminal groups on one side, and ordinary people spreading emotional or religious content through AI purely for monetization on the other. AI-related complaints have risen sharply, and the majority of victims in harassment and blackmail cases are women and children.
Are the Laws Working
Pakistan has cybercrime legislation in the form of PECA, the Prevention of Electronic Crimes Act. In theory, it provides a legal framework to pursue those creating harmful deepfake content. In practice, the law has more often been used against journalists, bloggers, and political opponents than against the actual creators of AI slop.
Nighat Dad believes that meaningful change will require consulting all stakeholders, including civil society, before creating new legislation. The risk of blunt laws is that they restrict legitimate speech while doing little to curb the actual harm being caused by AI fake videos in Pakistan.
The NCIA does contact social media platforms to request removal of deepfake content, but the scale of the problem far outpaces those efforts. Content is created faster than it can be reviewed, reported, or removed.
What Happens Next
Victor Rio raises a scenario that experts find deeply concerning. Right now, platforms pay Pakistani creators to produce AI content because it drives engagement. But the major social media companies are building their own AI video generation tools. Once those tools mature, the platforms will no longer need to pay outside creators. They will generate the content themselves.
At that point, millions of people who have built their livelihoods on creating AI fake videos in Pakistan and elsewhere will find their income gone. They will have an established infrastructure of fake accounts, AI skills, and audience reach. The concern is that they will use those resources in other ways, including fraud, coordinated scams, and targeted harassment campaigns. That would present an entirely different kind of challenge for law enforcement, one far more difficult to address than content moderation.
What Needs to Change
The responsibility for solving this problem sits in multiple places at once. Social media platforms created the incentive structure that made AI slop profitable, and they have the most direct power to change it by auditing monetized accounts and removing content that is demonstrably harmful or misleading before it spreads.
Law enforcement in Pakistan and elsewhere needs clearer legal frameworks that protect individuals from deepfake exploitation without being weaponized against free expression. That means inclusive legislation developed with civil society at the table, not rushed into law as a blunt instrument.
And for the many young Pakistanis who have entered this space simply to survive economically, the conversation needs to include genuine alternatives. Digital literacy, awareness of the real-world harm caused by disinformation, and access to ethical income streams in the digital economy are all part of a response that addresses root causes rather than symptoms.
AI fake videos in Pakistan did not emerge from nowhere. They emerged from poverty, limited education, algorithmic incentives, and a global content economy that values attention above all else. Understanding that is the first step toward changing it.
