AI Deepfake Defamation in Malaysia: What To Do If Fake Content Targets You

Gandhi Palanisamy
Jun 1
12 min read

Few things feel more violating than seeing a video of yourself saying something you never said. The face is yours. The voice sounds like yours. But the words, the scene, the whole moment were built by someone with an internet connection and a grudge. And by the time you find it, other people usually have too.

That is what AI deepfakes have made possible. Someone can now create a fake video, a cloned voice note, a doctored photo, a fake chat screenshot, or an entire fake account that makes it look as though you said or did something that never happened. It can surface on TikTok, Facebook, a WhatsApp group, or an anonymous page, and it can travel a lot faster than the truth.

If this has happened to you, your instinct is probably to hit back hard, or to explain yourself to everyone at once. Hold on a moment. What you do in the first day or two often decides whether this stays contained or blows up. So let us go through it calmly: what to save, what not to do, when a deepfake becomes defamation under Malaysian law, and when it is time to call a lawyer.

Key takeaways

Save everything first, before posts vanish: screen-record the full post, save the links, and note who shared it.
Stay calm. No angry replies, no public threats, and do not delete your own messages or records.
It may be defamation if the fake content is false, identifies you, was seen by others, and harms your reputation.
Four routes exist: platform takedown, a lawyer's letter of demand, a civil suit, and MCMC / police / online-safety law. Get advice early.

So what exactly is a deepfake?

A deepfake is any manipulated or AI-generated content, made to look real, that puts words, actions, or an image onto a person who never said or did the thing shown. In practice it takes a few shapes: a fake video of someone 'speaking', a cloned voice note, a doctored photo, an AI image built around a real person's face, a fake screenshot of a chat, or a fake account opened in someone's name. Very often it is a genuine clip stitched to a false caption, or a person's face used to sell a scam.

Not every edited clip is unlawful. A meme, an obvious parody, or fair comment may be treated very differently. The line is crossed when the content looks real, points to an identifiable person, and makes people believe something false and damaging about them. That is when the legal risk turns serious.

Can an AI deepfake be defamation in Malaysia?

Yes. AI deepfake defamation in Malaysia is treated like any other false and damaging publication: the fact that the content was generated by AI does not make it harmless in the eyes of the law. A fake video can damage a reputation just as a written post or a WhatsApp message can, and the courts treat it in much the same way.

To bring a civil defamation claim in Malaysia, you generally need to show three things. In plain terms:

It is defamatory. It makes ordinary, reasonable people think less of you.
It points to you. You are named, shown, tagged, or simply recognisable from the context.
Someone else saw it. It reached at least one person besides you.

What matters most is the meaning an ordinary viewer takes away, not the technical question of whether AI was used. If a fake clip makes people believe you committed a crime, cheated your customers, took a bribe, faked your qualifications, or did something shameful, that is the kind of message a court takes seriously. And the more damaging the false claim, the more urgent your response should be.

The caption can hurt you more than the video

Here is something a lot of people miss. Often it is not the video itself that does the damage. It is the words wrapped around it. The caption, the hashtags, the voiceover, the pinned comment, the reply underneath; that is frequently where the real sting lives.

A blurry or ambiguous clip can become defamatory the moment a caption tells people what to think: 'This lawyer forged documents.' 'This doctor takes bribes.' 'This contractor cheats his customers.' The image gives it a face; the caption supplies the accusation. That is exactly why, when you preserve evidence, you have to capture the whole post and not just the photo or video file.

First, and most important: save the evidence

If fake content about you is going around, move quickly, because posts get deleted, stories expire, and accounts disappear. Do not rely on your memory, and do not assume it will still be there tomorrow.

The single most useful thing you can do is screen-record the whole post while it is still live: the account, the caption, the comments, the date and time, the view and share counts, all in one capture. On top of that, save the link, save the account profile, and save any messages forwarding the content to you, especially the ones where other people react as though it is real. If it has cost you anything, a cancelled booking, a worried client, a lost enquiry, keep that too. And if you still have the original photo or clip that was twisted to make the fake, hold on to it.

In short: capture the content, capture who saw it, and capture the harm. That is the trio your lawyer will need.

What not to do

This is the part people most often get wrong in the heat of the moment, so it is worth slowing down for.

Do not reply in anger. A furious public post along the lines of 'I know who did this and I will destroy them' can create a second legal problem for you, and it warns the other side to start deleting evidence. Do not rally friends to swarm the account either, because the pile-on of comments and counter-attacks only makes the whole thing harder to control. Do not delete your own messages or records, even the awkward ones, because they often hold the context that helps your lawyer most. And be careful about reposting the fake just to deny it, since that can quietly hand it a bigger audience. If you do need to say something in public, keep it short and calm.

One more thing worth understanding early: a police report, an MCMC complaint, a platform report, and a civil claim are not the same thing. They do different jobs, and the right one depends on your situation. More on that next.

Your options, route by route

At a glance

Platform report: fastest removal, but no compensation and often no name behind the account.
Letter of demand: when you know the poster, can secure takedown, an apology, sometimes compensation.
Civil court action: for serious harm, damages plus a court order to stop it spreading.
MCMC / police / Online Safety Act: for scams, threats, obscene content, or anything involving a child.

The four legal routes after an AI deepfake in Malaysia: platform report, letter of demand, civil court action, and MCMC, police or the Online Safety Act.

Route 1: report it to the platform and ask for a takedown

Most platforms, TikTok, Facebook, Instagram, X, YouTube and the rest, have built-in tools to report impersonation, manipulated media, harassment, nudity, or scams. Choose the category that genuinely fits, because a deepfake scam advertisement is reported differently from a defamatory fake post, and a non-consensual intimate image is treated as urgent under its own rules. Keep a record of every report you file: the reference number, the date, and whatever the platform says back.

A takedown can stop the immediate bleeding. What it usually will not do is compensate you for the harm already done, or tell you who was behind the account. For that, you need the routes below.

Route 2: a lawyer's letter of demand

If the content is defamatory and you know, or can identify, who posted it, a letter of demand from your lawyer is often the next move. It can ask for the content to come down, a written apology or correction, an undertaking not to repeat it, and, depending on the facts, compensation.

A well-judged letter often resolves things quickly, because most people would rather take a post down than be sued over it. But it has to be done properly. A letter that is too aggressive, too vague, or legally shaky can backfire and turn a private dispute into a public quarrel, which is the last thing you want.

Route 3: civil court action

Where the harm is serious, you can sue. In a defamation suit you might seek damages for the injury to your reputation, additional (aggravated) damages where the other side acted maliciously, and, just as importantly, a court order, an injunction, to stop the content being published any further.

That last part matters most when something is spreading fast and money alone cannot undo the damage. In those cases your lawyer may apply to the court urgently. Whether suing is the right call depends on the strength of your evidence, who the wrongdoer is, how serious and how widespread the content is, and whether any judgment can realistically be enforced.

Route 4: MCMC, the police, and online-safety laws

Some fake content goes beyond a private reputation dispute. Where it involves a scam, obscene material, threats, harassment, anything concerning a child, or harm spreading across a licensed platform, criminal and regulatory routes can come into play alongside, or instead of, a civil claim.

Two laws come up often. Section 233 of the Communications and Multimedia Act 1998 deals with the improper use of network services, and is the provision frequently raised over offensive or menacing online content. The Online Safety Act 2025 (Act 866), which came into force on 1 January 2026, sets up a framework for reporting harmful content on licensed platforms and for action by the Malaysian Communications and Multimedia Commission (MCMC) in certain situations.

None of this means every nasty post is a crime, or that every deepfake falls under online-safety law. The real skill is classifying the problem correctly: is this mainly an injury to your reputation, a financial scam, an obscene image, harassment, something involving a child, or content racing across a major platform? The honest answer to that question is what decides whether your first move is a platform report, a police report, an MCMC complaint, a lawyer's letter, an urgent court application, or some combination of them.

What if the person behind it is anonymous?

Anonymous content is harder, but it is not always a dead end. Before the account vanishes, preserve every clue you can: the username and profile link, the profile photo, any connected or mutual accounts, fragments of an email or phone number, linked websites or payment details, a recognisable turn of phrase, even the watermark of the editing app. If you still hold the original file, its hidden data can matter too.

In some cases there are legal steps to uncover the person behind an account. Whether that is worth taking depends on how serious the content is, what records the platform keeps, the likely cost, and the realistic chance of getting something useful. A lawyer can give you a straight read on whether it is proportionate.

What if it is your business being targeted?

Businesses get hit by this too, and the damage can be every bit as real: a fake video of a director 'admitting' fraud, a doctored screenshot suggesting a company cheated its customers, a fake voice note racing through WhatsApp groups claiming a clinic, a contractor, or a firm is dishonest.

If your business is the target, preserve two things side by side: the content itself, and the commercial damage it caused, the drop in enquiries, the cancellations, the refund demands, the worried messages from staff or suppliers, the shift in your online reviews. The aim is not only to prove the content is fake. It is to show what it cost you, and what needs to happen now to stop it getting worse.

What if you are the one accused of posting or sharing it?

It can run the other way too. If you receive a letter of demand, a platform notice, or a call from the police over something you allegedly posted or shared, do not ignore it, but do not panic into admitting anything before you understand the allegation either.

Preserve everything: the content complained of, where you first received it, exactly what you posted or forwarded and when, and any context showing you treated it as comment, a warning, or a denial rather than a statement of fact. Then think honestly about the real question. Did you create the fake, or only pass it on? Did you believe it was real? Did your own caption add an accusation? Did you keep repeating it after you were warned?

This matters because, in defamation law, repeating a claim can land you in trouble of your own. Forwarding, reposting, stitching, duetting, even an added comment can count. So before you delete anything, apologise, or reply in public, get advice.

Does writing 'AI generated' protect you?

Not on its own. A label might cut down confusion in some cases, but it is not a shield. If the content still makes viewers believe something false and damaging about a real, identifiable person, the risk is still there, because a viewer may never read the label, and the caption may still carry the accusation. The question to ask yourself is a simple one: would an ordinary person seeing this come away believing a harmful claim about a real person? If the answer is yes, the label will not save it.

When is this urgent?

Speak to a lawyer quickly if the content accuses you of something serious, such as fraud, theft, corruption, abuse, sexual misconduct, fake qualifications, or a crime. The same goes if it uses your face, voice, name, or business identity, if it is being used for a scam, or if it involves intimate images or a child. Move fast, too, if it is spreading quickly, if customers, employers, or family have started asking you about it, or if you have already received a letter of demand or a police or platform notice and you are not sure how to respond.

The reason for the hurry is simple. With online content, the first day or two often decides whether the situation stays contained or runs away from you.

Before you see a lawyer, get this ready

You will get far more out of that first meeting if you walk in organised. Where you can, bring:

A short timeline of what happened.
Your screenshots and screen recordings.
The URLs and account links.
Whatever you know about the suspected person.
Anything that shows the content is fake.
Anything showing that people saw, believed, or shared it.
Evidence of the harm, financial or reputational.
Any platform report, police report, or legal letter you already have.

The more organised your evidence, the faster your lawyer can tell you whether to send a letter, push for a takedown, go to court urgently, or simply help you make a careful public statement.

How Gandhi Syahida & Associates can help

At Gandhi Syahida & Associates, we advise clients in Penang and across Malaysia on defamation, online harm, and social media disputes, from the first letter of demand through to civil litigation.

If AI-generated content has damaged your reputation, or you have been accused of posting or sharing it, the time to get advice is before you react in public. We can help you work out whether the content is defamatory, what evidence to lock down, whether a letter of demand makes sense, whether urgent takedown or court steps are worth considering, and how the civil side fits together with any platform, police, or MCMC route.

If it is already spreading, do not wait. Evidence disappears, accounts change, and screenshots lose their context. The sooner the strategy is set, the more can be done.

Frequently asked questions

Can I sue someone for an AI deepfake in Malaysia?

Often, yes. If the fake content is false, identifies you, was seen by others, and damages your reputation, it can support a defamation claim. The right route still depends on the content, the platform, your evidence, and who is behind it.

Is a fake AI video automatically defamation?

No. It has to carry a damaging meaning, point to an identifiable person, and have been seen by someone else. Genuine parody or obvious editing may be treated differently. But if it makes people believe something false and harmful about a real person, the risk rises quickly.

What should I do first if a deepfake of me is posted?

Save the evidence before anything else. Screen-record the full post with the account, caption, comments, and the date and time, and save the links. Then resist the urge to hit back or post a denial until you have had advice.

Can I report AI fake content to MCMC?

Depending on the facts, yes, particularly where it involves scams, obscene material, harassment, threats, a child, or misuse of network services. A lawyer can help you work out whether your situation is mainly civil defamation, a platform matter, criminal, regulatory, or a mix of these.

Can I claim compensation if a deepfake damaged my business?

Potentially. A business can suffer real reputational and financial harm from false content. Helpful evidence includes the content itself, proof that people saw it, customer messages, lost or cancelled work, and anything showing the claim was untrue.

I only forwarded it, I did not make it. Am I safe?

Not necessarily. In defamation, repeating a claim can carry its own risk, and forwarding or reposting counts as repeating. If you receive a legal letter over something you shared, preserve everything and get advice before you delete, apologise, or reply.

Should I publicly deny the deepfake?

Sometimes a careful clarification helps; sometimes it spreads the content further. The safer order is to preserve your evidence first, weigh the options, and keep any public statement short, factual, and calm.

Disclaimer

This article is general information, not legal advice. Every deepfake, defamation, and online-harm matter turns on its own facts, evidence, and stage, and on the law as it applies at the time. Please speak to a qualified lawyer before acting on anything here.