How to Report a Deepfake of Yourself on Social Media: 2026 Guide

How to report a deepfake of yourself to social media platforms depends on the harm: report the specific post and account through the category that matches it, and use a dedicated intimate-image process if the material is sexual. The detail that changes the response is not how realistic the clip looks: in the United States, covered platforms must remove qualifying non-consensual intimate images, including deepfakes, and known identical copies within 48 hours of a valid request, while ordinary impersonation complaints do not share that universal deadline (Federal Trade Commission [FTC], 2026).

A deepfake can falsely show you endorsing a product, use a cloned voice to ask family for money, impersonate your profile, or turn your likeness into a sexual image for humiliation or blackmail. The fastest route depends on the harm. An “AI-generated content” label is not proof of consent, and a misinformation report may not trigger the privacy process needed for an intimate image.

This guide treats reporting as an incident-response process rather than a list of buttons. It explains how to preserve enough evidence for a reviewer, which reporting route to choose on major platforms, how to limit search visibility and repeat uploads, and when to escalate beyond a platform. It also distinguishes removal of a post from removal of an account, deletion from a source website from suppression in a search engine, and a platform policy decision from a legal finding.

One global deepfake total would misrepresent a varied problem. Digital Rights Foundation reported 5,041 new Digital Security Helpline cases between May 2024 and December 2025 across several digital threats, including image-based abuse and AI-generated deepfakes; this is not a deepfake-only count (Digital Rights Foundation [DRF], 2026). The practical response should preserve evidence without increasing the audience for the abuse.

How to Report a Deepfake of Yourself to Social Media Platforms

Start at the original upload whenever possible. Open the post, video, image, profile or message thread that contains the deepfake, use the platform’s native Report control, and choose the most accurate category available. The wording differs by product and country; “impersonation”, “privacy”, “manipulated media”, “harassment”, “scam”, and “non-consensual intimate content” can lead to different review teams. The best category is the one describing what the content does, not simply the technology used to make it.

Next, report the account if it is pretending to be you, using your identity to contact others, or repeatedly publishing material. Do not assume that reporting a post will automatically remove the profile, or that reporting a profile will result in every post being reviewed. Treat each relevant object as a separate part of the incident: the original media, the impersonating profile, threats or messages, and any known re-uploads. Where the platform offers a privacy form in addition to the in-app report, submit it as well when your face, voice or personal information is being used without permission.

Write the report as a short factual statement. Identify yourself as the person depicted or the person whose voice is copied; state that you did not authorise the creation, publication or apparent endorsement; point to the exact section that is false or intimate; and explain the concrete risk, such as fraud, threats, impersonation or loss of control over intimate imagery. If the clip includes a false claim, describe the claim rather than arguing over whether the creator used a particular AI model. A reviewer needs to understand the violation and locate it quickly.

Avoid lengthy arguments, speculation about who created it, or accusations you cannot substantiate. If you know the account is impersonating you, say what makes that clear: the use of your name and image, a false claim to be your account, messages sent to contacts, or an invented endorsement. If you are uncertain whether the media is synthetic, you can still report a privacy violation or deceptive use of your identity; the report does not need to establish a forensic conclusion before the platform can review the conduct.

For a wider decision tree—especially when a deepfake also involves threats, fraud or a false public statement—see the magazine’s deepfake response sequence. Use that guidance as context, but follow the reporting interface and policies in force on the platform where the content is hosted.

Match the Report Category to the Harm

“Deepfake” is a description of how media may have been created or altered, not a universal moderation category. Platforms generally organise enforcement around harmful behaviour: impersonation, privacy violations, harassment, sexual exploitation, scams or deceptive manipulated media. A person who selects the wrong category may receive a refusal that addresses a different issue—for example, a response about AI disclosure when the actual complaint is unauthorised use of a real person’s likeness.

Choose the primary route from the harm table below, then add a second report if another violation exists. A fake account that posts a sexualised face-swap can involve impersonation and non-consensual intimate content. A voice clone used to request money can involve impersonation and fraud. Reporting the strongest, most specific violation first usually gives the reviewer a clearer basis for action than describing every problem as “fake news”.

What Happened	Best First Route	What to Explain
A profile claims to be you or acts as your official account.	Impersonation or fake identity.	State that you do not control or authorise the account; identify copied names, photos, bios or outreach.
A video, image or audio clip depicts you saying or doing something that never happened.	Manipulated media, deepfake, privacy or harassment, depending on the app.	Identify what was fabricated, where the relevant moment appears, and why the post misleads viewers.
The content falsely sells a product, solicits money or asks contacts for credentials.	Scam, fraud or deceptive identity, plus impersonation where available.	Describe the request, destination, financial risk and false affiliation.
A sexual image or video of you is fabricated or shared without consent.	Non-consensual intimate content or dedicated privacy/removal process.	State that the image is non-consensual and identify the exact post; avoid forwarding the image.
Someone threatens to publish a deepfake unless you pay or comply.	Threats, blackmail, sexual exploitation or extortion; report the account too.	Preserve the threat, deadline, payment demand and user details without engaging further.
A clip is misleading but uses parody, satire or commentary.	Context-specific manipulated media or impersonation review.	Explain the false impression and the absence of permission; acknowledge if the context is clearly labelled satire.

A label saying “AI-generated” does not settle consent, legality or harm. TikTok’s official guidance says it acts on content that violates its Community Guidelines regardless of whether it was altered with AI; YouTube likewise explains that its impersonation rules can apply when AI copies a voice or likeness to make a channel look authorised. In other words, disclosure can matter, but it is not a permission slip to impersonate someone or publish intimate imagery without consent (TikTok Support, n.d.; YouTube Help, n.d.).

Preserve Evidence Without Spreading the Deepfake

Evidence can disappear quickly when a post is deleted, an account changes its handle, or a story expires. Capture the minimum material needed to show what appeared, where it appeared, who posted it and when. Usually that means the exact post URL, account/profile URL, username and visible account ID, time and date, caption, relevant surrounding messages, the report confirmation and a short note of any consequences. Where the platform permits an export of messages or account data, retain it in its original format rather than relying only on cropped images.

Keep one private incident log. Record the date you discovered the content, the date it was published if visible, the URL, the account name at that time, the report category used, any case or ticket number, the response and your next action. If the creator edits the post, add a new entry rather than replacing the original note. Keep original files and screenshots in a secure folder with access restricted to people helping with the incident. If the matter may become a legal case, avoid editing source files and ask a qualified professional how to preserve the chain of custody.

Do not repost the deepfake to expose the creator or crowdsource an identity. Resharing can create more copies and widen its audience. For a non-explicit fake, a screenshot with the username and context may be enough. For nude or sexually explicit imagery involving anyone who was under 18 when it was created, do not download, duplicate or send the image as proof; preserve URLs and account details, then use the platform’s child-safety route and the relevant authority.

A common mistake is to upload a private or intimate image to a random “AI detector” website in the hope of proving it is fake. That may expose the file to another service, create further copies, or violate the service’s terms. Many detectors return a probability rather than a finding about consent, identity or authorship. For general verification tasks, the magazine’s seven checks for suspicious media explain why source history, context and corroboration should be considered alongside technical clues. This reporting guide does not require a detector score.

Preserve messages around the content, not just the image itself. A threat that says “pay by Friday or I will send this to your employer” can be more important for an extortion investigation than the deepfake alone. Save the sender, time, demand and any payment address or account reference. Do not reply in anger or threaten the poster back. If there is an immediate threat of physical harm, stalking or imminent exposure, prioritise your safety and contact an appropriate local support service or emergency authority rather than delaying to build a perfect evidence file.

Platform Reporting Paths: Instagram, Facebook, TikTok, YouTube and X

The available labels and help pages change over time, and some reporting options vary by region or account type. The table below summarises official routes that were available when checked for this article on 29 October 2026. It is not a guarantee that each menu will look identical on every phone. If a label is missing, use the platform’s Help Centre and search for the specific policy term rather than submitting repeated reports through unrelated categories.

Platform	First Report Route	When to Add a Second Route
Instagram, Facebook and Threads	Use the post’s or profile’s in-app Report control. Choose the closest option for impersonation, harassment, privacy, scam or non-consensual intimate content. Meta’s Safety Centre links to reporting resources for intimate-image abuse across its apps.	Report the fake profile separately when it is posing as you; use Meta’s dedicated intimate-image resources when sexual imagery or threats are involved.
TikTok	Open the post, tap Share or press and hold it, select Report, then choose Misinformation followed by Deepfakes, synthetic media, and manipulated media. Submit the report.	Report the profile separately if it impersonates you, and report threats or scam messages through the relevant safety route.
YouTube	Report the channel if it is impersonating you. If a video or your personal information uses your likeness or voice without consent, use the privacy-removal process described in YouTube Help.	Report the specific video where appropriate; a channel report and a privacy complaint address different issues.
X	Use the post-level Report control and select the closest policy route. X’s non-consensual nudity policy explicitly covers AI-generated or manipulated intimate imagery; impersonation has a separate route.	Report the account for impersonation and submit the privacy or intimate-content form for the specific media. The policy says an X account is not required for an impersonation report.
Any platform with a private support form	Use the platform’s Help Centre when the in-app category does not accurately describe a likeness, privacy or intimate-image violation. Include the exact URL and a concise explanation.	Keep the form submission ID and attach only what the platform specifically requests and what is safe and lawful to share.

TikTok’s official report path is Misinformation → Deepfakes, synthetic media, and manipulated media. YouTube distinguishes channel impersonation from privacy removal for realistic AI-generated likeness or voice. X treats impersonation and non-consensual intimate content as separate routes, and its intimate-media policy covers AI-generated or edited imagery (TikTok Support, n.d.; YouTube Help, n.d.; X Help Center, 2026).

When reporting for another adult, check representative eligibility; for a child, use the child-safety route. Do not impersonate a victim or publicly tag platform staff with intimate content.

When the Deepfake Is Sexual: Use the Dedicated Removal Processes

Non-consensual intimate imagery needs a different response from a fake endorsement or ordinary edited clip. Start with the platform’s explicit intimate-content or privacy route, not only a general deepfake or misinformation category. If the person depicted was under 18, treat it as a child-safety incident and do not download or forward the image.

Two free hash-based services address a narrower problem: matching eligible intimate images on participating services. StopNCII is for adults who meet its age requirements, including being 18 or older when the image was taken and over 18 now. NCMEC’s Take It Down is for nude or sexually explicit images taken when the person was under 18. Both generate a digital fingerprint, or hash, rather than send the original image; neither deletes every online copy.

Service	Who It Is For	How It Works and What It Cannot Promise
StopNCII.org	Eligible adults shown in intimate content, with the image still in their possession and the age requirements met.	Creates a hash on the user’s device and shares the fingerprint with participating companies. It can help limit future sharing on those services; it does not scan or remove every website, encrypted message or offline copy.
NCMEC Take It Down	People whose nude, partially nude or sexually explicit image/video was taken when they were under 18, including people who are now adults.	Creates a hash without uploading the image. Participating public or unencrypted services may match it; the tool cannot guarantee removal from every website or private channel.
Platform privacy / intimate-content report	Anyone who can identify a specific image or post hosted on a platform with a reporting process.	Targets a known post or content item and allows the service to review a policy or legal request. It may need to be repeated for other posts, accounts or services.

In May 2026, FTC Chairman Andrew N. Ferguson said, “In the age of AI, anyone can be targeted, and that becomes even more appalling if children are involved” (FTC, 2026). The US Take It Down Act requires covered platforms to remove qualifying non-consensual intimate imagery and known identical copies within 48 hours of a valid request. This is not a deadline for every deepfake or country.

If a covered platform lacks a removal process or fails to act, the FTC accepts complaints. Keep the request and timestamp; this route enforces platform compliance but does not erase copies on every third-party or encrypted service.

Use only an image lawfully in your possession and follow eligibility rules; never download a circulating copy just to create a hash. For images created when you were under 18, the process is designed to keep the file on your device. The official pages for StopNCII and Take It Down explain their limits.

Reduce Search Visibility and Track Copies

Removing a social post does not necessarily remove the file from the wider web. The same deepfake may have been reposted by another account, copied to a forum, included in a search result, or saved by people who downloaded it before moderation. Treat each location as a separate target. Start with the earliest known source, then record the additional URLs where the content appears and send each hosting service its own report. A removal decision on one platform is evidence for a follow-up report elsewhere, but it does not automatically bind another service.

Search engines create a second layer of visibility. If an intimate image, private identifying information or a harmful result appears in search, use the search engine’s official personal-content removal request where eligible. A search engine may remove a result from its search listings without removing the content from the website that hosts it. That distinction matters: people may still reach a page through a direct link, another search engine or a private message. The source website’s own removal process remains important.

Track copies in a simple log with one row per URL. Record the domain, post or image URL, account, date discovered, which form you used, the case number and the outcome. Avoid building a public spreadsheet that repeats the deepfake itself or exposes intimate details. A private log makes it easier to identify recurring accounts, distinguish a new upload from a duplicate link and show a reviewer that the problem is continuing after an earlier report.

Hash-based matching is useful only within the participating services and file types supported by that service. Editing, cropping, changing the aspect ratio, adding text or audio, or generating a new variation can mean that a different file is not treated as an identical copy. Conversely, the absence of an exact match does not prove that a new upload is legitimate. Report newly created variations as new content, and describe the connection to the earlier case without claiming the platform can automatically recognise every modified version.

The technical side of the problem is developing quickly. The magazine’s review of how realistic deepfakes have become explains why visual glitches are not a dependable threshold for action. You do not need to establish that a clip is a perfect synthetic artefact before reporting a deceptive use of your likeness. The essential facts are the content’s location, what it purports to show, whether you authorised it, and what harm or risk follows from its publication.

Escalate When the First Report Is Rejected or Ignored

A refusal is not always the end of the process. Read the reason carefully and identify the issue the moderator considered. If the response says there is no impersonation because the account has a disclosure, but the content is a fabricated intimate image, submit the appropriate privacy or non-consensual-intimate-content report. If a profile report was accepted but the video remains online, report the specific video. If the system cannot locate the post, check whether the URL is correct, the content is still public, or the author has changed the username.

Appeal through the platform’s documented review or privacy process rather than sending the same generic text repeatedly. Keep the appeal concise: refer to the original case number, state what the first decision appears to have missed, identify the platform rule or removal category that matches the content, and provide the exact post URL. Add new evidence only when it clarifies the harm—for example, a follow-up threat, a second impersonation account or a re-upload after removal. Avoid uploading a large volume of unrelated screenshots, which can make the relevant facts harder to review.

Escalate outside the platform when the incident involves blackmail, extortion, stalking, threats, financial fraud, a child, or serious intimate-image abuse. The correct authority depends on where you and the accused person are located and what happened. A police or cybercrime report can create an independent record, while a victim-support organisation can help with safe reporting and practical next steps. Platform reports and law-enforcement reports serve different purposes; one is not a substitute for the other.

For US incidents covered by the Take It Down Act, the FTC has a platform-compliance route if a valid intimate-image request is not handled as required. For other types of harm, identify the appropriate consumer-fraud, privacy, data-protection or police process in your jurisdiction. Do not assume that an overseas platform’s reporting policy is the same as the criminal law where you live. When money, employment, immigration, custody or physical safety is at stake, consider qualified legal advice instead of relying only on community posts or generic AI guidance.

If evidence may be used in a legal dispute, preserve the original file where you already possess it lawfully, retain metadata where possible, note how you obtained it and avoid repeatedly converting or compressing it. Do not try to perform your own invasive investigation, hack an account or trick the suspected creator into sending more material. For a deeper discussion of what courts may need to assess, see the magazine’s explanation of the evidential value of deepfake audio. That discussion is not a substitute for advice from a lawyer or forensic expert who understands the relevant jurisdiction.

Know Which Legal Route Applies in Your Country

Platform policies and local laws are different systems. A service may remove content without a criminal offence being established, while unlawful content may remain online after a moderation decision. Legal routes depend on the content, the person’s age, intent and jurisdiction; the US deadline is not universal.

In the United States, the Take It Down Act requires covered services to accept valid requests for qualifying non-consensual intimate imagery, including deepfakes, and remove the content and known identical copies within 48 hours. The FTC accepts complaints about covered platforms that fail to provide a process or respond as required. The law does not guarantee removal of every non-sexual fake or copy outside its scope.

In June 2026, Canada’s Privacy Commissioner found that X Corp. and xAI violated federal privacy law by launching Grok’s image-generation tool without adequate safeguards. Commissioner Philippe Dufresne said, “Organizations have a responsibility and legal obligation to protect Canadians’ fundamental right to privacy” (Office of the Privacy Commissioner of Canada, 2026). The finding is specific to Canadian law and this case, but illustrates that privacy routes may be separate from platform moderation.

In Pakistan, people facing online impersonation, blackmail or image-based abuse can contact the National Cyber Crime Investigation Agency or seek survivor support from a digital-rights organisation. DRF reported 5,041 helpline cases across several digital threats between May 2024 and December 2025—not a deepfake-only count—and said 92% of surveyed complainants reported reduced risk after support. Founder Nighat Dad called deepfakes and related abuse “tools of control against women and marginalised communities in Pakistan”. She added, “Digital violence is not only virtual. It is actively reshaping people’s safety, careers, and ability to participate in public life” (DRF, 2026).

Where You Are	Useful Escalation Route	Important Limit
United States	For qualifying non-consensual intimate imagery, use the platform’s legal notice process; if it fails, consult the FTC’s Take It Down enforcement guidance.	The 48-hour requirement applies to covered intimate-image requests, not all deepfakes.
Canada	Consider the applicable privacy regulator and local police where there are threats, coercion or other potential crimes.	A regulator’s finding in a specific case does not automatically decide every separate complaint.
Pakistan	Consider the National Cyber Crime Investigation Agency for cyber harassment and a survivor-support organisation such as Digital Rights Foundation.	Report routes and laws can change; verify current official instructions and keep a case record.
Other jurisdictions	Contact the local police or cybercrime service when a crime or safety risk may be involved, alongside the platform’s reporting path.	Do not assume a US or UK rule, deadline or remedy applies where you live.

Match the route to its purpose: platforms moderate hosted content, search engines review listings, privacy regulators assess data practices, and law enforcement investigates possible offences. Keep records portable; one institution may not be able to order another service to remove material.

Understand What Happens After a Report

Once a report is submitted, the platform may remove the content, limit its visibility, label it, ask for more information, decline action, or take action against the account. It may not tell you every enforcement detail, particularly where another user’s account is involved. A report confirmation means that the submission was received; it does not mean a reviewer has confirmed the content is fake or that removal is guaranteed. Read the outcome message and note whether it refers to the post, profile, message or policy issue you actually reported.

If a post is removed, check the original URL again and update your log. Then search the account’s visible activity, without engaging with it, for other copies or related posts. If a known mirror remains, report that exact location. If the same person creates a new account, report the new profile and mention the earlier case if the platform allows it. Avoid using automated scripts or creating multiple accounts to flood moderators; mass reporting can impair review and may violate platform rules.

If the decision is “no violation found”, separate two questions: is the item actually synthetic, and does its use violate the selected policy? A genuine photo may be used deceptively with a false caption; a synthetic clip may be harmless satire; and a highly realistic deepfake may be clearly disclosed but still violate a privacy or impersonation rule. Explain the actual effect. When the issue is a false public claim, source context and corroboration may matter more than visual artefacts. The magazine’s guide on how to verify a public video claim is designed around that distinction.

After reporting, verify the original URL, report known re-uploads separately, warn contacts or customers if needed, and secure any potentially compromised account. If harm continues, move from repeated generic reports to the documented appeal, privacy, legal-notice or local-authority route.

A delay does not prove that no action was taken, but keep your own record. Review times vary by policy, region, language and volume. The FTC’s 48-hour provision is limited to covered US intimate-image requests; it does not promise a similar outcome for a voice clone, fake profile or general manipulated-media complaint. A quick removal does not establish who created the content.

Protect Your Accounts and Limit Follow-On Harm

A deepfake often becomes more damaging when it is paired with an access attempt, false endorsement, payment request or harassment campaign. If the creator may have accessed your account, change the password from a trusted device, enable multi-factor authentication, review active sessions and connected apps, and secure the email account used for password recovery. If a fake profile is messaging your contacts, warn them through a known official channel and ask them not to send money, identity documents, verification codes or private images in response to the fake.

For a cloned voice, tell family members, colleagues or staff to confirm unusual requests through a separate channel that they already trust. Calling back the number displayed by a suspicious caller is not enough if that number could be spoofed or controlled by the attacker. A pre-agreed phrase can support a family plan, but should not be treated as the sole security control. Businesses should require a second approval for payment changes, payroll instructions and sensitive account recovery rather than relying on a person’s voice or face alone. For a practical explanation of voice-clone risks in phone calls, see the magazine’s separate technical guide.

If someone demands money, more images or silence, do not assume payment will stop the abuse. Preserve the demand and report it through the applicable extortion or exploitation route. Avoid negotiating alone if doing so increases your risk. Tell a trusted person and seek specialist support, especially if the person making threats knows your address, workplace, school or family. Do not publish their private details in retaliation; doxxing can create new risks and complicate an investigation.

Choose a proportionate public response. Sometimes a brief notice from an authenticated account—“This profile is not mine; do not send it money”—can prevent confusion. Sometimes silence is safer while a platform or authority handles a case. You do not owe the public a forensic demonstration of why the video is fake, and you should not be pressured to share the intimate content again in order to defend yourself. When a false video concerns employment, a client relationship or public office, use an official communications contact and a documented correction route rather than a long argument in the comments.

Public-facing professionals can keep a record of official accounts, trusted contact channels and an escalation contact. That preparation makes impersonation easier to document and helps colleagues verify unusual requests before acting.

Our Editorial Verification Process

This guide was prepared on 29 October 2026 by comparing platform-level reporting instructions with primary sources from TikTok Support, YouTube Help, Meta’s Safety Centre and X Help Center. The response pathways were checked for differences between impersonation, manipulated media, privacy complaints and non-consensual intimate imagery. The legal and escalation discussion was cross-checked against the Federal Trade Commission’s 2026 enforcement announcement, the Office of the Privacy Commissioner of Canada’s June 2026 release, and Digital Rights Foundation’s May 2026 research release.

The analysis also reviewed the official descriptions of StopNCII and NCMEC’s Take It Down to distinguish eligibility, local hash generation and the limitations of participating-platform matching. The 48-hour deadline is attributed only to covered non-consensual intimate-image requests under the US Take It Down Act. DRF’s 5,041 cases are described as a broad digital-security caseload rather than a deepfake count; its 92% figure is attributed to surveyed complainants, not the entire population.

No live reports were submitted on behalf of a victim, no personal intimate files were uploaded, and no forensic detector or platform moderation system was experimentally benchmarked for this article. Menu labels can vary by region and app version, and the available reporting procedures may change after publication. Readers should check the linked official help page before sending a report and should seek local legal or emergency support when the incident involves threats, coercion or immediate danger.

AI assistance disclosure: This article was researched and drafted with AI assistance. The sources linked in the reference list were checked against official materials where available, but final human editorial review and publication sign-off should be completed by the publisher before publication. No unperformed firsthand testing or independent human review is claimed in this draft.

Conclusion

Reporting a deepfake works best when the response is matched to the harm. Identify the exact content, report the post and any impersonating profile separately, preserve the evidence that can be shared safely, and use the specific privacy or intimate-image route when the issue involves sexual imagery. For a fake voice or false endorsement, add the fraud, impersonation or privacy category that accurately describes the risk. If an appeal is needed, refer back to the first case and explain the policy issue the initial decision missed.

Removal is rarely a single action. The source post, search listing, re-upload, private message, fake account and threatening contact may each require a separate report. Hash services can help with certain intimate images on participating platforms, but they do not erase every copy. The US Take It Down Act’s 48-hour requirement is meaningful for qualifying requests on covered services, yet it should not be presented as a worldwide guarantee for every kind of manipulated content.

The most important distinction is between preserving evidence and amplifying abuse. Keep the URLs, account details, timestamps, report records and threats; do not circulate intimate imagery to make a case. Escalate serious threats or sexual exploitation to the appropriate local authority, and use specialist support when the process becomes overwhelming. Platform interfaces and legal routes will continue to change, so the safest response is specific, documented, and grounded in the current official policy rather than assumptions about how deepfake moderation ought to work.

Frequently Asked Questions

How Do I Report a Deepfake of Myself to Social Media Platforms?

Report the exact post under the closest category—deepfake, manipulated media, impersonation, privacy or non-consensual intimate content. Report a fake account separately, save the case number and use the dedicated intimate-image route for sexual content.

Can I Report a Deepfake If I Do Not Have an Account on That Platform?

Sometimes. X accepts impersonation reports without an account, and some services offer external privacy forms. Availability varies by platform and category, so check the relevant official help page.

Should I Prove That the Video Was Made With AI Before Reporting It?

Not always. Describe the unauthorised or deceptive use of your identity. A forensic finding may help in a serious dispute, but a detector score is not required for every report and does not prove consent or authorship.

Does an AI-Generated Label Make a Deepfake Acceptable?

No. Labels improve transparency but do not permit impersonation, fraud, harassment or non-consensual intimate imagery. Platforms can apply their rules regardless of AI disclosure.

Can StopNCII Remove Every Copy of an Intimate Deepfake?

No. StopNCII hashes eligible images on the user’s device and shares the fingerprint with participating companies. It cannot guarantee removal from every website, encrypted conversation or private device.

What If the Deepfake Depicts Me When I Was Under 18?

Do not download or forward the image. Preserve the URL and account details, use the platform’s child-safety route, and consider NCMEC’s Take It Down for qualifying intimate images created when you were under 18.

What Should I Do If a Platform Rejects My Report?

Read the reason, then appeal through the correct privacy, impersonation or intimate-image route. Keep the case number and report ongoing threats or extortion to the relevant local authority.

Will Reporting One Post Remove the Whole Account and Every Copy?

Not necessarily. Posts, profiles, messages, search results and re-uploads may be handled separately. Report each known location and use a separate search-removal or eligible hash-matching route where appropriate.