AI-generated nude images are becoming increasingly realistic and increasingly easy to create. A photograph taken from social media can potentially be manipulated within minutes to make it appear that a person is nude or engaged in sexual activity, even though no genuine intimate photograph ever existed.
Singapore law has now been strengthened specifically to address this problem. The Penal Code definition of an intimate image expressly includes certain material that has been altered or generated to appear to depict a person, and the law now criminalises certain non-consensual production of intimate images. The amendments addressing synthetic AI-generated intimate material came into force on 17 August 2026.
For victims, however, the immediate concern is often broader than whether a criminal offence has been committed. They may need the image removed, further distribution stopped, an anonymous perpetrator identified, threats addressed, and evidence preserved before it disappears.
GJC Law has extensive experience advising and representing clients in harassment and related matters.
Our lawyers understand that cases involving intimate material can require both an immediate response and a longer-term legal strategy. Depending on the circumstances, the available options may involve the Police, the Protection from Harassment Act (POHA), the Online Safety Commission and civil proceedings.
Key Points
- A deepfake nude can potentially be illegal even though no genuine nude photograph ever existed.
- Singapore law now expressly covers certain AI-generated and manipulated intimate images.
- Producing a non-consensual intimate image can itself constitute an offence in specified circumstances, distribution is not always required.
- Possessing, accessing, distributing or threatening to distribute intimate images may also constitute offences where the statutory requirements are met.
- POHA may provide additional remedies where deepfakes are used to harass, humiliate or distress a person, and if a protection order is needed.
- Since June 2026, victims of certain online harms can also seek assistance through Singapore’s Online Safety Commission.
- Victims should preserve the original evidence, URLs, usernames, messages and account information before seeking removal.
Table of Contents
What Is a Deepfake?
Are Deepfake Nudes Illegal?
When Are Deepfakes Unlawful?
What Can Victims Do?
POHA Protection Orders
Online Safety Commission
Evidence to Preserve
Common Misconceptions
When to Speak to a Lawyer
How GJC Law Can Help
What Is a Deepfake?
A Deepfake is an AI-generated or AI-manipulated image, video or audio made to appear as though a real person said or did something they did not.
In the context of sexual or intimate deepfakes, this may involve:
- placing someone’s face onto a nude body;
- digitally removing clothing;
- depicting someone apparently engaged in sexual activity; or
- generating an entirely synthetic sexual image or video that appears to depict a real person.
An important legal point is that the body or sexual activity depicted does not necessarily have to be genuine.
The Penal Code’s definition of an intimate image or recording includes material that has been altered or generated to appear to depict the relevant person in specified intimate circumstances, unless it has been altered or generated in such a way that no reasonable person would believe it depicts that person. The legislation itself gives the example of someone’s face being pasted onto an image of another person’s body engaged in a sexual act.
This means that saying “it isn’t really her body” does not necessarily take an AI-generated nude outside the law.
Is Creating a Deepfake Nude Illegal in Singapore?
It can be.
Singapore has introduced an offence covering the production of an intimate image or recording without the depicted person’s consent where the person producing it knows or has reason to believe that its production will or is likely to cause that person humiliation, alarm or distress.
The provision expressly recognises that material can be produced by generating, altering or manipulating it. The offence is punishable by imprisonment for up to two years, a fine, or both, with enhanced consequences where the depicted person is below 14 years old.
This is significant because the legal issue no longer begins only when somebody uploads or forwards the image.
The production itself may potentially constitute an offence if the statutory elements are satisfied.
When Can a Deepfake Become Unlawful?
There is no single “deepfake offence” covering every possible use of artificial intelligence. The applicable law depends on what was created, what the person did with it and why.
Sexual or Intimate Deepfakes
Sections 377BD and 377BE of the Penal Code are particularly relevant.
Depending on the facts, the law can address the non-consensual production, possession or access, distribution, or threatened distribution of intimate images or recordings.
Section 377BE, for example, criminalises certain intentional or knowing distribution of an intimate image without the depicted person’s consent where the offender knows or has reason to believe the distribution will or is likely to cause humiliation, alarm or distress. It also covers certain threats to distribute such material.
A person therefore should not assume that sending an image privately to one other person avoids potential criminal liability.
Deepfakes Used to Harass or Humiliate Someone
The Protection from Harassment Act (POHA) may also be relevant.
Sections 3 and 4 now expressly refer to sexual or indecent words, behaviour or communications in the context of harassment, alarm, distress or humiliation.
Depending on the circumstances, a sexually humiliating deepfake sent directly to a victim, circulated among colleagues or classmates, or posted online could therefore raise issues under POHA as well as the Penal Code.
Deepfakes Used to Demand Money or Force Someone to Act
A perpetrator might threaten:
“Pay me or I will send this video to your family.”
Or:
“Do what I say or I will publish these images.”
Depending upon the precise threat and demand, this may potentially engage offences relating to intimate images as well as extortion or other criminal offences.
Importantly, for the offence of threatening to distribute an intimate image under section 377BE, it is not necessary for the prosecution to prove that the threatened image actually exists.
Deepfakes Used to Obtain Money by Deception
Deepfake technology is not limited to sexual images.
AI-generated video or audio might impersonate a company director, relative, customer or business counterparty and instruct someone to transfer money.
Depending on the circumstances, such conduct could potentially amount to cheating or other fraud-related offences.
Deepfakes Making False Allegations
A manipulated recording could also falsely portray somebody saying or doing something damaging to their reputation.
Depending upon what was represented and how it was published, defamation may become relevant. POHA also provides remedies concerning qualifying false statements of fact. The Singapore Courts recognise several forms of false-statement orders, including stop publication and correction orders.
Deepfakes Involving Businesses or Brands
AI-generated material may also falsely suggest that a business, professional or public figure endorses a product or is associated with a particular company.
Depending upon how names, trade marks, branding, images or copyrighted material are used, issues involving trade mark infringement, passing off, copyright or other intellectual property rights may arise.
The legal analysis will depend on precisely how the material was created and used.
What Can a Victim of a Deepfake Nude Do?
The appropriate response depends on whether the immediate priority is stopping distribution, identifying the perpetrator, obtaining protection, pursuing criminal action or seeking compensation.
Preserve the Evidence
Before blocking accounts or requesting removal, preserve the evidence wherever it is safe to do so.
The important point is to preserve information capable of establishing what occurred, where it appeared, when it appeared and who may have been responsible.
Make a Police Report
Where a potentially criminal intimate deepfake has been created, possessed, distributed or threatened, the victim can consider making a police report.
Useful information may include:
- the original URL;
- username and account details;
- dates and times;
- relevant messages or chats;
- threats or demands;
- details of suspected perpetrators; and
- information about people who received the material.
A cropped screenshot alone may omit information that could later help identify the account or trace what happened.
Consider a Protection Order Under POHA
Where the conduct amounts to harassment, a victim may potentially seek a Protection Order (PO) under POHA.
Where urgent intervention is required, an Expedited Protection Order (EPO) may also be available. An EPO provides temporary protection while the substantive Protection Order application is being determined.
A Protection Order can potentially require the respondent to stop the harassment and remove a publication causing harassment.
Where false statements are involved, POHA also provides separate false-statement remedies in appropriate cases.
Seek Help from the Online Safety Commission
There is now another important avenue for victims.
The Online Safety Commission (OSC) commenced operations on 29 June 2026 under the Online Safety (Relief and Accountability) Act 2025 (OSRAA).
Its initial areas include intimate image abuse, image-based child abuse, online harassment, doxxing and online stalking. The Commissioner can, where the legal requirements are met, direct action such as removing harmful material or restricting a perpetrator’s account. The framework can also assist in obtaining identifying information about anonymous users for specified purposes.
OSRAA also introduced statutory torts that can, in applicable cases, allow victims to pursue remedies including damages and injunctions against responsible online actors.
This can be particularly important with deepfakes because stopping an image from continuing to circulate may be just as urgent as establishing who originally created it.
Listen to Noelle Teoh Discuss Online Safety on CNA938
What Evidence Should a Victim Preserve?
Where possible, preserve:
- full screenshots showing the account name, date and time;
- URLs;
- screen recordings showing the relevant page or account;
- original Telegram, WhatsApp, social-media or other messages;
- threats or demands accompanying the material;
- records showing who received the image or video;
- information identifying the suspected perpetrator;
- evidence of money demanded or transferred; and
- correspondence with platforms concerning takedown requests.
Preserve the evidence carefully and consider obtaining advice about the appropriate way to provide it to the Police, a platform or your lawyers.
Common Misconceptions About Deepfake Nudes
“I generated it but never posted it.”
That does not necessarily mean no offence has been committed. Singapore law now expressly criminalises certain non-consensual production of intimate images, including AI-generated material.
“It isn’t her real body.”
That is not necessarily a defence. An altered or generated image can fall within the statutory definition where it appears to show the person in the specified intimate circumstances and a reasonable person could believe it depicts them.
“I only sent it to one friend.”
Distribution does not necessarily require posting something publicly or sending it to a large audience. Whether an offence is made out depends on the statutory requirements and circumstances.
“I only threatened to release it but did not do so.”
Threatening to distribute an intimate image can itself constitute an offence where the requirements of section 377BE are satisfied. Significantly, the prosecution does not necessarily have to prove that the threatened image actually exists.
When Should You Speak to a Lawyer?
Deepfake cases can move quickly. Once material has been shared through messaging apps, social media accounts or online groups, identifying where it originated and limiting further circulation can become increasingly difficult.
Legal advice may be particularly useful where:
- the perpetrator is known to you;
- intimate material has already been distributed;
- you are being threatened or blackmailed;
- the perpetrator is anonymous;
- material continues to be reposted;
- the platform has not removed it;
- you require urgent protection; or
- significant reputational or financial harm has occurred.
The appropriate response may involve more than one legal route at the same time.
How GJC Law Can Help You
GJC Law has extensive experience in contentious matters involving harassment and harmful online conduct. Where intimate or deepfake material is involved, our lawyers can assess the circumstances and advise on the legal options available, including Police reports, POHA proceedings, urgent protective relief and remedies available under Singapore’s newer online-safety framework.
Where the matter also involves threats, extortion, intimate-image offences or other potential criminal conduct, our criminal defence and litigation team can advise on the criminal-law implications and the appropriate next steps.
If you have discovered that a deepfake image or video depicting you has been created, circulated or threatened, early advice can be particularly important before evidence disappears or the material spreads further.
Contact GJC Law to arrange a consultation and discuss the legal options available in your circumstances.
