AI-generated and digitally altered sexual images are giving old forms of harassment a faster, cheaper distribution system. Cases reported in India, Bangladesh and Pakistan show how a fabricated image can trigger humiliation, blackmail, political intimidation, family retaliation and, in the most serious allegations, physical danger. They do not establish a precise regional “flood” or prevalence rate: comparable South Asian data remain limited. They do establish that the harm is real, gendered and difficult to contain once copies move across platforms and borders.
What counts as a deepfake or synthetic intimate abuse?
A deepfake is AI-generated or AI-altered audio, image or video that falsely depicts a real person. In abuse cases, the technology can be sophisticated—or as simple as placing a face on another body.
As an Amazon Associate I earn from qualifying purchases.
- Non-consensual intimate imagery (NCII): real intimate material shared without permission, plus synthetic sexual imagery using someone’s likeness.
- Face-swapping and morphing: a person’s face is digitally attached to another image or video.
- Synthetic impersonation: fabricated speech, sexual conversations or actions attributed to someone who never said or did them.
- Gendered disinformation: false material intended to sexualize, shame, discredit or silence someone because of gender.
India’s 2025 Ministry of Electronics and Information Technology FAQ specifically includes AI-generated sexual deepfakes of identifiable people, synthetic undressing, morphed nudity, face-swapped intimate imagery and voice-cloned sexual harassment. The examples are set out in the ministry’s official FAQ.
Cases from India, Bangladesh and Pakistan show different kinds of danger
The documented cases below are illustrations of harm, not a statistically complete count of victims or files.
#1 Best Overall
| Case | What was reported | Why it matters |
|---|---|---|
| Rashmika Mandanna and other Indian actors | A fake video placed Mandanna’s face on another woman’s body. The 2023 regional report also said Katrina Kaif, Alia Bhatt and Deepika Padukone had been targeted. | Celebrity visibility can expose the abuse quickly, but public attention does not prevent reputational damage or replication. |
| Bangladesh politicians Rumin Farhana and Nipun Roy | Manipulated sexualized videos circulated before the January 7, 2024 election. | Election-period disinformation can intimidate candidates and spread faster than verification. |
| An 18-year-old woman in Kohistan, Pakistan | Police said a doctored photograph circulated before the woman was allegedly killed by relatives in an honour-related attack. | The image’s role and the precise chain of causation were allegations reported by police, not a final judicial finding. The case shows why online abuse can become a physical-safety issue. |
| Shahzadi Rai, a transgender Karachi municipal councillor | Rai described abusive trolling and deepfake images as threats to her safety and career. | For transgender public figures, sexual humiliation can combine with outing, misgendering and political intimidation. |
The cases were reported by Scroll in a Thomson Reuters Foundation investigation published on December 16, 2023: regional case reporting.
Why the harm falls hardest on women and gender-diverse people
Sexual reputation is used as a weapon
In patriarchal and conservative settings, a false suggestion of sexual conduct can affect family relationships, housing, employment and political credibility even when viewers eventually learn it was fabricated. People may remember the allegation while forgetting the correction.
Private individuals have fewer ways to fight back
Public figures may have lawyers, media access and direct contacts at platforms. A private person may face the same viral abuse with no institutional escalation route, while also being more exposed to family or community retaliation.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Queer and transgender targets face layered attacks
A synthetic image can be used to out someone, misgender them, threaten conversion-related abuse or discredit an activist. Risks vary by country, family and community; there is no single experience shared by all LGBTQ+ people.
“Obviously fake” still can be socially effective
Realism is not the same as technical invisibility. A crude edit can cause harm if it is believable enough for a recipient, arrives in a trusted messaging group or is repeated by many accounts. Pakistani digital-rights advocate Nighat Dad told Scroll that even people who recognize a fake may still see the target’s integrity and credibility damaged.
Generative tools have lowered the cost of abuse
Image generators, face-swapping services, “undressing” tools and voice-cloning systems allow perpetrators to produce and revise material without advanced specialist skills. Public photographs and a social-media account may be enough to start. Multiple languages, captions and versions can be generated quickly, then distributed through public networks, closed groups and encrypted messaging.
Detection systems are imperfect. A detector’s result should not be treated as conclusive proof that an image is real or fake, particularly across languages, skin tones, genders and regional media formats. The practical threat is the combination of cheap creation, rapid copying and social credibility.
Free tools Windows power users keep installed
One-click scans. No signup required.
What the available evidence can—and cannot—say about scale
There is no comparable public dataset that measures South Asia by victims, generated files, posts, reports, takedowns or criminal investigations. Platform transparency reports generally do not provide consistent country, gender, sexual-orientation and synthetic-content breakdowns.
The 2023 reporting repeated an expert estimate that more than 90% of deepfake videos online are pornographic, but it did not provide separate South Asia data. That figure should not be presented as a current regional statistic. Increased visibility may reflect both more abuse and better public recognition of it; underreporting is also likely, but its extent is not established here.
Rank #4
The damage continues after an image is debunked
- Personal safety: panic, stalking, threats, sextortion and possible family or community violence.
- Work and income: lost employment, clients or political opportunities, plus legal, counselling and repeated takedown costs.
- Democratic participation: sexualized disinformation can discourage women and queer people from standing for office or speaking publicly.
- Persistent circulation: deleting one post does not remove copies, screenshots, search results, private messages or overseas-hosted files.
- Unequal enforcement: moderation may work differently by language and region, and systems can wrongly suppress consensual LGBTQ+ expression while missing abusive synthetic content.
What India’s current rules require
India’s legal position has changed since the 2023 reporting. Under the consolidated Information Technology Rules published in 2026, when an individual complains about material showing nudity or sexual acts, exposing private areas or involving electronic impersonation—including artificially morphed images—an intermediary must take reasonable and practicable measures to remove or disable access within two hours. The rules also address unlawful synthetic information, including non-consensual intimate imagery, sexually explicit deepfakes, deceptive impersonation and false depictions of a person or event. Synthetic material not covered by those prohibitions must be prominently labelled and, where technically feasible, carry provenance or metadata signals.
The obligation depends on the applicable intermediary category and complaint route. It is not a promise that every deepfake will disappear everywhere within two hours. Copies, search listings, closed groups and foreign-hosted material may remain. The relevant texts are the consolidated 2026 IT Rules and the February 2026 MeitY notification. MeitY also published an NCII Standard Operating Procedure in October 2025.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallCrashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minutePlatform remedies are useful but separate
Google Search and Google services
Google says an identifiable person can request removal of fake sexual or nude imagery that depicts them in an intimate state without consent. Search removal can be requested through Google Search’s sexual-image process. In the current image viewer, open the result, select More, choose Remove result and identify it as a sexual image of you; the interface distinguishes real NCII from fake AI imagery. Labels can vary by country, account and product version. Search de-indexing does not delete the source page or file.
Best Value
- Book - 1, 000 books to read before you die: a life-changing list (1000 before you die)
- Language: english
- Binding: hardcover
Google provides a separate process for removal from services such as Drive and Photos through Google Account Help. YouTube reports use a separate NCII reporting route. An authorized representative may submit a request for someone who cannot safely do so themselves.
What to do if you or someone you support is targeted
- Check immediate safety first. If there are threats, stalking or a risk of family violence, contact local emergency services and a trusted person or support organization before pursuing takedowns.
- Do not amplify the material. Do not forward it to prove the abuse or quote-search distinctive captions unnecessarily.
- Preserve essential evidence safely. Record URLs, usernames, timestamps, platform names and threats. Take screenshots only where safe and lawful. Never download or circulate explicit material involving anyone under 18.
- Report the original post and major copies. Keep confirmation numbers and dates for each report.
- Request search removal separately. Removing a result from Google does not remove the host file.
- Use a representative where possible. This can reduce repeated exposure to the intimate material.
- Get local legal and digital-rights help. In Pakistan, Google’s support guidance refers users to the Digital Rights Foundation’s cyber-harassment helpline. Elsewhere, contact a local digital-rights, domestic-violence or LGBTQ+ support organization.
- For minors, use child-safety reporting channels immediately. Do not retain, send or repost the material.
What remains unresolved
Fast removal can limit spread but may rely on automated decisions and incomplete evidence. Demanding extensive proof delays help and shifts the burden onto victims. Platforms may remove an image while leaving threats, captions or coordinated accounts online. Perpetrators can move to encrypted groups, and police may treat a case as a reputation dispute rather than harassment, extortion, privacy invasion or gender-based violence.
The central question is therefore not simply whether a viewer can spot a fake. It is whether the target can regain safety, privacy and public credibility after a fabricated sexual image has entered circulation.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
