YouTube introduces a deepfake protection system, but many MENA journalists say the move comes too late as they continue facing escalating digital harassment and misinformation.
Amira Khaled got a call last year from a friend in Cairo who'd seen a video of her. She was giving a speech, but Amira hadn't given that speech. She'd never even worn that outfit. A deepfake. A convincing one. It circulated on WhatsApp groups for hours before she even knew it existed. By the time she could get it taken down, it had reached thousands of people in Egypt, and damage to her credibility was already done. Stories like Amira's are becoming routine in MENA newsrooms.
Now YouTube says it has a solution. The company announced Tuesday that it's expanding its likeness detection technology to politicians, government officials, and journalists in a pilot program. Members of this group can now request the removal of AI-generated deepfakes that impersonate them. It's a meaningful step. It's also three years too late for the journalists and civic leaders already living in a kind of digital warfare most people in the West haven't had to experience.
The technology itself has been available to YouTube's creator program since last year, but it operated quietly in the background. Now YouTube is bringing it to people whose deepfakes could actually swing elections, start conflicts, or destroy careers. To use the tool, participants must upload a selfie and government ID. They can then flag content impersonating them and request removal if it violates YouTube policy. Not all requests will succeed. YouTube says it will evaluate whether content is parody or legitimate political critique, which are protected forms of speech.
This is where the initiative's limitations become apparent. In MENA countries like Egypt, Saudi Arabia, Pakistan, and the UAE, the line between legitimate political critique and incitement is often blurred by the government. What one person sees as parody, another sees as dangerous misinformation. YouTube's algorithm will struggle with these distinctions. When a journalist in Riyadh flags a deepfake, will YouTube remove it? When a Palestinian journalist in East Jerusalem flags a deepfake, will it be treated the same way as one flagged by a Western journalist? These questions matter enormously for people living in regions where digital speech is already highly monitored and controlled.
The problem is also that YouTube's solution addresses the symptom, not the disease. Deepfakes are emerging in MENA at an accelerating pace, and most of them target journalists, activists, and opposition figures. According to researchers who study disinformation in the region, deepfake attacks have increased by roughly 60 percent in the past year alone. A Pakistani journalist told TechScoop that she's had at least three deepfakes created in her image in the past eighteen months. An Egyptian television host had a deepfake used to spread false claims about his health. These aren't isolated incidents. They're part of a coordinated effort to undermine trust in media and destabilize public discourse in countries where journalists already face intense pressure.
The real issue is that bad actors have learned something YouTube is only now beginning to address: deepfakes work best when they target people who are already marginalized or controversial. A deepfake of a Western politician makes headlines. A deepfake of a journalist in Cairo goes quiet because the person it's attacking doesn't have the same platform or resources to respond. YouTube's tool requires verification through government ID, which assumes people have faith in that ID system. In some MENA countries, that's not a safe assumption for vulnerable journalists.
Leslie Miller, YouTube's Vice President of Government Affairs, framed the expansion as a matter of "integrity of the public conversation." That's true, but the conversation YouTube is protecting is primarily the one happening in developed democracies. The conversation happening in MENA newsrooms, where journalists are contending with deepfakes, surveillance, and governmental pressure simultaneously, is operating under completely different rules.
YouTube also announced it's supporting the NO FAKES Act in the U.S. Congress, which would regulate unauthorized AI recreations of voices and likenesses. This is good policy. But it applies only to the United States. A journalist in Beirut or Karachi can't rely on U.S. legislation to protect her from a deepfake created by someone in a different country and distributed globally. The infrastructure for detection and removal has to be built at a global scale, and YouTube's current approach is regional at best.
What would actually help is more aggressive detection before content goes live, the ability for journalists in vulnerable regions to verify their identity through alternative means, and explicit protections for people in countries where freedom of speech is already constrained. YouTube says it plans to eventually prevent violating content from going live in the first place. That's the feature that matters. Until then, the tool is largely reactive: a way to close the barn door after deepfakes have already stampeded through it.
For journalists like Amira, watching YouTube announce this tool now feels bittersweet. It would have helped last year. It might help protect her next year. But the deepfakes that have already done damage to her reputation and career can't be undone by a feature that comes years late. The real challenge for YouTube—and for the wider tech industry—is recognizing that threats to journalists and civic leaders aren't evenly distributed. They're concentrated in regions where media is already under siege. Solutions that work for everyone require understanding that reality first.
