AI Deepfakes and Celebrity Videos: What Rashmika Mandanna's Case Teaches Us

When a Fake Moment Looks Real

You open Instagram and see a Bollywood actress in a video. She appears to be making a controversial statement, promoting a product or meeting someone.

The face looks familiar. The voice sounds convincing. Nothing immediately seems unusual.

But what if the actress was never there?

Artificial intelligence can now generate realistic images, imitate voices and change existing videos. Some of these manipulated recordings are known as deepfakes. The technology itself has many legitimate uses, including filmmaking and creative projects. The concern begins when fabricated content is presented as a real event. A false video can become a trending story before anyone checks whether the event actually happened.

And the person shown in it may suddenly find themselves answering questions about something they never did.

What Happened to Rashmika Mandanna?

One of Bollywood's most widely discussed deepfake controversies involved actress Rashmika Mandanna in November 2023. A video circulating online appeared to show Rashmika entering an elevator. But there was a problem. The woman in the original recording was not Rashmika. The original video featured British-Indian social media personality Zara Patel. Someone had digitally replaced her face with Rashmika's, creating the impression that the actress appeared in the recording.

Fact-checking helped expose the manipulation. According to BBC News, Rashmika described the experience as frightening and expressed concern about the wider misuse of technology. Zara Patel also expressed distress over the manipulation of her original video. The incident raised an important question: if a familiar celebrity can be convincingly inserted into someone else's recording, how easily could an ordinary person face the same problem?

What happened after the video went viral?

The controversy did not end with online criticism. In January 2024, Delhi Police arrested a man accused of creating and circulating the manipulated video. According to The Indian Express, police alleged that the accused created the deepfake to increase followers on a social media fan page. The arrest was a significant development, although an arrest should not be confused with a final court finding of guilt.

This case showed how a digitally manipulated video could move from a social media post to a police investigation. It also highlighted something easily overlooked: two different women were affected. One person's face was used without consent, while another person's original recording was altered.

The Problem Isn't Limited to Celebrity Gossip

Imagine seeing a video of your favourite actor recommending an expensive investment or announcing a giveaway. Because the face and voice appear familiar, the offer might seem trustworthy. But a celebrity appearing in an advertisement is not proof that they agreed to promote the product. The US Federal Trade Commission has warned about fake celebrity endorsements, including manipulated audio and video used by scammers.

The same problem can affect entertainment news. A fabricated interview can make an actor appear to express an opinion they never shared. A false photograph showing two stars together can start relationship rumours. A manipulated clip can create controversy around an event that never occurred. In each situation, viewers may form an opinion about a real person based on something invented.

Not every AI-generated image is misleading. Some are clearly labelled as fiction, parody or creative art. The important difference is whether viewers are being led to believe the content documents a real event.

Why Sharing a Fake Video Makes Things Worse

The person who creates misleading content is only one part of the story. Another person downloads it. A fan page reposts it. Someone shares it in a WhatsApp group. Soon, the video reaches people who have never seen its original source. Some users may knowingly spread misinformation. Others may genuinely believe the content is authentic. Their intentions are different, but both can contribute to its circulation.

Even after a video is exposed as fake, copies may remain online. People who watched the original might never encounter the correction. This is especially troubling when the content involves a person's reputation, privacy or safety. And there is no automatic rule that everyone who shares manipulated content has broken the law. Legal responsibility depends on the circumstances and applicable laws.

Still, sharing responsibly matters even when no legal issue is involved.

How to Check Whether a Celebrity Photo or Video Is Real

You don't need advanced technical skills to question suspicious content.

Eroszooming recommends starting with three simple checks.

1. Find the original source

Ask where the image or video first appeared. Was it uploaded by the celebrity's verified account? Did it come from an original interview, film promotion or public appearance? Or is it circulating through unfamiliar pages without any explanation? A verified account can provide useful context, but it isn't absolute proof that everything posted is accurate.

2. Look for the full context

A short clip may leave out important details, even when the footage itself is genuine. Search the celebrity's name along with a few words describing the claim. Look for official statements, the complete interview or independent reporting. If the claim is significant but no reliable source confirms it, avoid treating it as established fact.

3. Check the image's earlier history

For a photograph, try Google Lens. You can upload an image or screenshot to search for similar pictures and earlier appearances online. Sometimes, an image presented as a new celebrity controversy is actually an old photograph with a misleading caption. A reverse image search may uncover that earlier context. However, finding no matching image does not prove that a photograph is genuine.

Can AI detection tools tell the truth?

AI detectors may offer clues, but they can produce incorrect results. Some files also contain Content Credentials, which can describe how an image was produced or edited. They provide useful information about a file's history, not a guarantee that the event shown is true. So the safest approach combines source checking, context and independent confirmation.

If the evidence is incomplete, the right answer is sometimes simply: We don't know whether this is real yet.

What If Someone Uses Your Face Without Permission?

This problem doesn't only affect film stars. A student, teacher, office worker or social media user could also become the subject of a manipulated image or recording. If you discover realistic fake content using your identity, there are several sensible steps to consider. First, preserve evidence. Save the link, account name, date and relevant screenshots before the content disappears. You don't need to spread the image further to document what happened.

Second, report the content. Use the platform's reporting tools for impersonation, privacy violations or misleading synthetic media, as appropriate. Third, request removal. Some platforms have processes specifically for realistic manipulated content. For example, YouTube's privacy guidelines allow eligible people to request removal of realistic AI-generated or altered content resembling their face or voice. Requests are reviewed rather than automatically approved.

Finally, seek appropriate support. If the content involves threats, blackmail, intimate imagery, harassment or financial fraud, contact relevant local authorities or a qualified legal professional. The exact options depend on where you live and the circumstances.

Are Social Media Platforms Doing Enough?

Platforms are trying to improve the way they identify misleading synthetic content. YouTube, for example, expanded access to its likeness-detection technology for entertainment industry participants in April 2026. The system helps eligible participants find AI-generated content resembling them and request its removal. But technology alone cannot solve the problem.

Labels may be missing. Manipulated content can travel between platforms. And detecting that AI was used does not always explain whether a post is misleading. There is also a responsibility on publishers. Entertainment pages and news websites should avoid presenting unverified photographs, manipulated interviews or AI-generated celebrity moments as genuine news. When something is fictional or illustrative, clearly identifying it helps protect both the people depicted and the audience.

Why This Matters Beyond Bollywood

Consider what happened in Rashmika Mandanna's case. One woman's original video was altered to show another woman's face. The manipulated version circulated widely enough to attract public attention and a police investigation. Now imagine a similar situation involving someone who isn't famous. They may not have millions of followers, a management team or the resources to respond publicly.

A fabricated recording could create problems at work, damage a relationship or cause serious personal distress. This is why AI impersonation is more than a celebrity news issue. It's about consent, trust and a person's ability to control how their identity is represented. AI can be useful and creative. But people deserve to know when a scene is fictional, when a recording has been manipulated and when an apparent statement was never made.

What Happens Next?

As AI tools improve, identifying manipulated content by appearance alone may become increasingly difficult. Better verification tools, clearer labels, responsible platform policies and legal protections will all have roles to play. But viewers can already make one useful change. The next time a surprising celebrity image or video appears on your feed, check where it came from before reacting or sharing.

Because sometimes the face is real, but the moment never happened.