Deepfakes, Misinformation, and Trust

A convincing video once circulated online showing a well-known public figure saying something they never actually said. It was entirely generated by AI, and to an untrained eye, it was indistinguishable from real footage.

What a Deepfake Actually Is

A deepfake is a piece of synthetic media, an image, video, or audio clip, generated or altered by AI to realistically depict someone doing or saying something that did not actually happen. The same generative technology discussed in Chapter 4 that can produce a beautiful piece of digital art can, in the wrong hands, produce a fabricated video of a real person.

The technology has improved so quickly that the old advice to ‘look for glitches’, unnatural blinking, strange lighting, blurred edges, is no longer reliable on its own. The most current, high-quality deepfakes can pass casual visual inspection entirely.

Why This Matters Beyond the Obvious

The immediate danger of deepfakes is being fooled by a specific fake video or audio clip. The deeper danger, sometimes called the liar’s dividend, is what happens to trust in general: once people know convincing fakes exist, genuine footage of real events can also be dismissed as fake by anyone who finds it inconvenient. The mere existence of deepfake technology can be used to cast doubt on real evidence, which is its own form of damage, separate from any single fake being believed.

This makes media literacy, the practice of critically evaluating what you see and hear before accepting or sharing it, more important now than at any previous point, not less, even as the fakes themselves become harder to detect by eye.

Beyond Video: Voice and Audio Fakes

Deepfakes are often discussed as a video problem, but AI-generated voice cloning has become one of the most practically dangerous forms of synthetic media, precisely because it requires so little source material, sometimes just a few seconds of someone’s voice from a social media video, and because a phone call feels inherently more trustworthy and urgent than a text message. Scams impersonating a family member’s voice in a fabricated emergency, or impersonating a manager’s voice to authorise an urgent payment, have already caused real financial and emotional harm.

The same core defence applies here as with video: a convincing voice is no longer, on its own, proof that the person on the other end is who they claim to be. Establishing a family code word for genuine emergencies, or simply calling the person back on their known number before acting on an urgent request, are small habits that meaningfully close this gap.

Practical Habits for an Uncertain Media Landscape

Before sharing anything emotionally charged, especially content designed to provoke immediate outrage or fear, pause and check whether reputable, independent news sources are reporting the same thing. A single unverified clip, however convincing, is not evidence on its own. Check who originally posted the content and whether the account has any history or credibility. Reverse image searches and searching for a video’s key claim in quotation marks can often reveal whether a piece of content has already been identified and debunked elsewhere.

Most importantly, build the habit of treating strong emotional reactions, outrage, fear, or triumphant vindication, as a signal to slow down and verify, rather than a reason to share immediately. Misinformation, AI-generated or otherwise, is specifically designed to spread fastest when it triggers exactly those emotions.

An Older Instruction Than the Internet

None of this is new to a Muslim. Long before a video could be faked, the Qur’an gave the believer the same instruction for any report brought by someone whose word cannot be relied upon:

The word the verse uses is fatabayyanu: make it clear, establish it, before you act. A forwarded clip from an unknown account is exactly the kind of report the verse describes. The habit this chapter has been building, to pause, to check, and only then to share, is that verse practised on a phone.

In a world where seeing is no longer automatically believing, the most valuable skill is not detecting fakes by eye, but building the calm, consistent habit of checking before trusting and before sharing.