Fake It till You Break It: Inside the Deepfake Era

Previously, I believed that if one could see something, it could be trusted because seeing is believing. A video was proof. A voice recording was evidence. That belief is gone, and I don’t think we have fully sat with the implications of that. There is an abnormal shift happening. AI-generated images, voice cloning, face swapping apps, and similar tools were supposed to expand how people express themselves, but instead they have slowly chipped away at what mattered the most: the belief that people can express themselves through their images and their words.

I attended a workshop under the project titled “Digital Trust and Free Expression in the AI Era” at Department of Media and Development Communication, University of the Punjab, supported by the U.S. Mission to Pakistan in partnership with the Pakistan-U.S. Alumni Network (PUAN) last week. During the discussions on digital trust. I took part in the discussion and shared my thoughts that we are now in a time where we must take this out of the realm of fun tech novelty and place it in the realm of a structural issue in how societies operate. Everyone says deepfakes are a “misinformation” issue, but this is not an adequate way to describe it. Misinformation is something that’s a false statement that can, in theory, be fact-checked. Deepfakes strike at the heart of video and audio credibility which have always been treated as vivid reality. When people realize that any clip can be fabricated, they stop trusting all clips, even real ones. This is what concerns me most. Fake things are believed and real things become deniable. Consider the implications of journalism, courts, political accountability, etc. If a politician is recorded saying something damaging, they can simply claim that it’s a “deepfake” and, depending on how much trust citizen have lost, people may believe them.

We are creating a society in which it is possible to negotiate authenticity and in which negotiable authenticity is a gift for those who wish to avoid consequences. I also face a conflict here: I strongly believe in free speech, satire, parody, art and provocation even with synthetic media. The same technology behind de-aged actors, resurrected voices, dubbing, and experimental AI is extremely creative, yet it is all connected to malicious deepfakes. There is no way to ‘clean’ a tool of its worst use-case. A hammer can break windows or build houses. I do not think banning synthetic media is the answer, nor do I support pre-emptive censorship. That path harms legitimate satire, journalism and art under the pretense of ‘preserving truth’. History does not look kindly on institutions that decide what is a “fair” fake. Freedom of expression has never been of consequence or friction, and right now, deepfakes have almost no friction.

The reality is there is no cost, disclosure, and accountability in using this technology deceptively. The issue is the imbalance. I believe the solution is not reducing synthetic tools but improving provenance. We need a digital watermarking, which cannot be “easily removed”: cryptographic signatures embedded into cameras, microphones and generative tools at the point of creation, so that the “unverified” label appears by default, not as an afterthought. Not to prohibit any content but to allow people to know what they are seeing before deciding whether to trust it.

Platforms have been allowed to operate without responsibility for too long. Convincing fakes can reach millions before correction and corrections never travel as far. This asymmetry is not natural; it is a design choice. Platforms can mark, rate or tag synthetic content at the time of dissemination. Most simply haven’t had to! But, beyond platforms, we must change our own instincts as individuals.

The old saying was “pics or it didn’t happen.” The new rule should be: “Who gains from believing this and can I verify it myself? A society that believes everything or worse, believes nothing is in danger. The deepfake age is not the age of AI; it is the consequences of a world where the cost of faking reality is near zero and the cost of establishing truth remains high. Freedom of expression should protect the most challenging and profound exchange of ideas, not the production of fake consent, fake evidence, or fake people. We can have creative freedom and still expect honesty about sources. These are not competing prioritie

These are not different objectives/priorities. However, if we pretend it’s a content-moderation footnote, we must make sure we are not waking up in a world where “fake it till you make it” has become ‘fake it till you break it” and by that time, it won’t be just one video or one lie that will break. It will be our fundamental ability to reach a shared understanding of reality.

The views, thoughts, and opinions expressed in this blog are solely those of the author and do not necessarily reflect the official policy or position of the U.S. Mission to Pakistan or USEFP.

The writer, Mahnoor Ashraf Wattoo, is a student of Media and Development Communication at University of the Punjab, Lahore