What platforms require of AI-generated content

Rights to what you make are the tool's business. But it is the platform that takes a post down, and its rules are its own. The spread is wider than people think: some require a mark at upload and penalise its absence, some do not mention AI in their rules at all, and some simply forbid generated material.

Below is what the platforms' own rules say. The rules of this topic change more often than most.

Facebook

Is a label required. Required for photorealistic video and realistic-sounding audio. Word for word: "We require people to disclose, using our AI-disclosure tool, whenever they post organic content with photorealistic video or realistic-sounding audio that was digitally created or altered, and we may apply penalties if they fail to do so." There is no such requirement for images: the platform promises to label them itself The rule is common to Facebook, Instagram and Threads — Meta keeps it in the "Misinformation" section of the "Community Standards". The platform's own help article ("Label AI content on Instagram", help.instagram.com/761121959519495) is unavailable to an ordinary request: the domain returns HTTP 400 without a browser, and the robots.txt of help.instagram.com contains "User-agent: ClaudeBot / Disallow: /" source, checked 2026-08-04

How it is labelled. The label is called "AI info" (previously "Made with AI"): "Meta began adding labels to a wider range of video, audio and image content" (since May 2024), "These are displayed as “AI Info”". The creator uses a separate tool: "we’re adding a feature for people to disclose when they share AI-generated video or audio so we can add a label to it". The platform shows the origin of the label: "shares whether the content is labeled because of industry-shared signals or because someone self-disclosed" The wording about the disclosure tool is from the Meta blog https://about.fb.com/news/2024/02/labeling-ai-generated-images-on-facebook-instagram-and-threads/. There is no step-by-step instruction (where exactly the toggle is in the posting form) in the available sources: Meta's help pages are closed, see the note to ai_disclosure_required source, checked 2026-08-04

Does the platform detect it itself. Yes, through industry metadata. Word for word: "we’re building industry-leading tools that can identify invisible markers at scale – specifically, the “AI generated” information in the C2PA and IPTC technical standards"; "in the coming months we will label images that users post to Facebook, Instagram and Threads when we can detect industry standard indicators that they are AI-generated". Also: "Our “Made with AI” labels on AI-generated video, audio and images will be based on our detection of industry-shared signals of AI images or people self-disclosing that they’re uploading AI-generated content" Both standards are named directly — C2PA and IPTC. The second quotation is from https://about.fb.com/news/2024/04/metas-approach-to-labeling-ai-generated-content-and-manipulated-media/. Meta admits the count is incomplete: "labeling methodology is still evolving" and the data "may not capture labels on some content that is edited with AI" source, checked 2026-08-04

What is forbidden. There is no direct ban on AI content — there is a label and a ban on harm. Word for word: "We may place an informative label on the face of content – or reject content submitted as an advertisement – when the content is a photorealistic image or video, or realistic sounding audio, that was digitally created or altered and creates a particularly high risk of materially deceiving the public on a matter of public importance." Removed regardless of the way it was produced: "We remove misinformation or unverifiable rumors that expert partners have determined are likely to directly contribute to a risk of imminent violence or physical harm to people"; "We consult with leading health organizations to identify health misinformation likely to directly contribute to imminent harm to public health and safety"; on elections and the census — "We remove misinformation that is likely to directly contribute to a risk of interference with people’s ability to participate in those processes" Meta has no list of "you may not make voices of real people / faces / elections / medical claims" separately for AI: the prohibitions are worded by harm, not by the way of production. Hence — partial source, checked 2026-08-04

What happens if you break the rule. "…and we may apply penalties if they fail to do so" — the only wording about punishment for a missing label. What exactly the punishment is (removal, limited distribution, loss of monetization) the rule does not disclose No terms, steps or thresholds are named in the rule — only "may apply penalties". No more detailed page about the consequences of a missing label was found on the available Meta resources source, checked 2026-08-04

Can you monetise it. AI is not addressed directly; the rule is worded about unoriginal content: "Content that is duplicative or involves minor edits to another creator’s post" is considered "unoriginal and deprioritized"; "Creators who are not adding substantial creative value will see their content deprioritized, and they may see their account deemed non-recommendable and demonetized, if they continue to post primarily unoriginal content" The word "AI" is not in the text of this Meta blog entry about the originality of content — that is, the platform does not punish AI as such, it punishes unoriginality. The "Content Monetization Policies" themselves could not be checked: they are at www.facebook.com/business/help, a domain closed in robots.txt ("User-agent: * / Disallow: /") source, checked 2026-08-04

Platform card

Instagram

Platform card

How it is labelled. The label is applied by the platform, not by the creator: "LinkedIn members will start seeing a new icon that helps members trace the origin of AI-created media, including the source and history of the content, and whether it was created or edited by AI" The official news of the LinkedIn press centre of 15 May 2024 ("LinkedIn adopts C2PA standard"). No tick box at upload for the creator is described on the available pages. The help article "Content credentials" is at www.linkedin.com/help/linkedin/answer/a6282984 — the domain is closed, see the note to ai_disclosure_required source, checked 2026-08-04

Does the platform detect it itself. Yes, by the C2PA standard: "Today we shared in a post from Patrick Corrigan, our VP of Legal, how we are rolling out the new C2PA AI-generated content standard on the LinkedIn platform"; the icon shows "the source and history of the content, and whether it was created or edited by AI" The press centre news.linkedin.com is open (its robots.txt is empty), unlike the main domain. There are no mentions of IPTC or of its own detectors on the available pages source, checked 2026-08-04

What is forbidden. The prohibition is worded through harm, not through the way of production. According to the search index of the official rules page, LinkedIn removes "Synthetic or manipulated media, such as doctored images or videos that distort real-life events and likely to cause harm to the subject, other individuals or groups, or society at large"; the general rule is "Do not share content that is false, misleading, or intended to deceive" The domain www.linkedin.com is closed to us in robots.txt: separate groups there list "User-agent: anthropic-ai", "ClaudeBot", "Claude-Web", "Claude-User", "Claude-SearchBot" — each with "Disallow: /". LinkedIn's community guidelines and help centre are only on this domain, they may not be read and the prohibition must not be bypassed. The words are given according to the search index for this same official page; their exactness has not been re-checked against the page itself — hence partial source, checked 2026-08-04

Platform card

Is a label required. The creator has no obligation to label AI content — the platform applies the label itself. An indication by the creator is called possible but not obligatory: "Some Pins may also have the AI modified label if the content owner indicated that the Pin was generated or modified using AI". The Pinterest "AI acceptable use guidelines" contain no labelling requirements at all Three official pages were checked: the help article about the labels (help.pinterest.com/en/article/gen-ai-labels), the community guidelines (policy.pinterest.com/en/community-guidelines) and the AI use guidelines (policy.pinterest.com/en/genai-acceptable-use-guidelines). None of them has the words "must label". Read through WebFetch: the robots.txt of both domains disallows "anthropic-ai", "ClaudeBot", "Claude-Web", but not "Claude-User" source, checked 2026-08-04

How it is labelled. The label is called "AI modified" and appears from the platform itself: the user sees it "in close-up at the bottom left after you click on the Pin". For advertising pins the disclosure is under the ellipsis icon, with the wording "was modified with AI" or "may be modified with AI". The creator may dispute the label: to argue about it one has to "contact our support team" Neither a tick box at upload nor a field "this was made by AI" in the pin creation form is described in the help centre source, checked 2026-08-04

Does the platform detect it itself. Yes, in two ways: "Pinterest analyzes a Pin’s metadata following the IPTC Metadata Standard, but also developed classifiers that automatically detect Gen AI content, even if the content doesn’t have obvious markers" It is the IPTC standard that is named (not C2PA). Plus its own detectors, which work without metadata. The same is in the platform's news (newsroom.pinterest.com/news/introducing-gen-ai-labels/) source, checked 2026-08-04

What is forbidden. The general one: "Our policies apply to everyone and to all types of content, including synthetically generated content". Prohibited: "Content that is synthetically generated or manipulated to degrade and/or shame" (bullying); "synthetically generated or manipulated content that impersonates a person or organization" — for this "We’ll deactivate accounts"; on elections — "Fabricated or manipulated content (e.g. content generated with AI tools) making someone appear to do or say something they didn’t with the clear intent of influencing public sentiment about a political figure and/or election"; "Instructions, tools or services for creating sexually explicit or nudity content (e.g. “nudifiers”)" Pinterest's ban regarding elections is worded directly and with a direct mention of AI tools. There is no separate ban on medical claims made by AI — there is a general ban on medical misinformation, not tied to the way of production source, checked 2026-08-04

Platform card

What is forbidden. There is no separate ban on AI content. The closest is rule 5: "Be authentic. You don’t have to use your real name, but do not intentionally mislead others or impersonate an individual or entity in a deceptive manner"; and rule 2 about "content manipulation" The detailed page of the rule about impersonating someone else (support.reddithelp.com/hc/articles/41180423371156) is closed by Cloudflare (HTTP 403), so it was not possible to check whether it mentions deepfakes. The Reddit rules were read in full at redditinc.com/policies/reddit-rules (the robots.txt of this domain allows the /policies section to everyone). The detailed explanations of the rules are at support.reddithelp.com — there is Cloudflare protection there, HTTP 403 and "Just a moment... Enable JavaScript and cookies to continue"; www.reddit.com is closed by the rule "User-agent: * / Disallow: /". It must not be bypassed. Note also that WebFetch is not allowed onto redditinc.com ("Claude Code is unable to fetch from redditinc.com"), the page was obtained by an ordinary GET. source, checked 2026-08-04

Platform card

Snapchat

Is a label required. For an ordinary post there is no obligation; for a monetized one there is. Word for word from the monetization rules: "Content that leverages sophisticated editing or AI-based tools is monetizable—if it is original, entertaining, or informative, not misleading, and if the use of AI is disclosed anywhere in the content or creator profile." The "Community Guidelines" contain no requirement to label AI, there is only the general: "We also expect Snapchatters to use AI responsibly" Disclosure is allowed "anywhere in the content or creator profile" — that is, anywhere in the post itself or in the account description, no separate field is required. The community guidelines were checked at https://values.snap.com/privacy/transparency/community-guidelines source, checked 2026-08-04

How it is labelled. No separate tick box at upload is described in the rules: the disclosure is made in one's own words — "disclosed anywhere in the content or creator profile". For AI content made with Snapchat's own tools the label is applied by the platform: "AI-generated content that was created within Snapchat is eligible for recommendation. Our tools always include transparency indicators" What the "transparency indicators" look like on shots made with Snapchat's tools is not described on the pages that were checked source, checked 2026-08-04

What is forbidden. The community guidelines: "We prohibit spreading false information that causes harm or is malicious, such as denying the existence of tragic events, unsubstantiated medical claims, undermining the integrity of civic processes, or manipulating content for false or misleading purposes (whether through generative AI or through deceptive editing)"; bullying — "sharing humiliating imagery (including manipulated or AI-generated imagery)"; the general one — "We reserve the right to take appropriate enforcement action against accounts that use AI to violate our Community Guidelines or abuse our AI products. This can include termination of an account". The monetization rules separately prohibit "Deceptively manipulated media, including content that has been generated or enhanced with the use of AI tools to manipulate users (for example, editing an image of someone’s body or face to imply a radical transformation…)" Medical claims ("unsubstantiated medical claims") and the undermining of the "integrity of civic processes" (that is, elections) are named directly alongside what is made by AI. The last quotation is from the page https://values.snap.com/policy/creator-monetization-policy source, checked 2026-08-04

What happens if you break the rule. For violating the standards with the help of AI: "We reserve the right to take appropriate enforcement action against accounts that use AI to violate our Community Guidelines or abuse our AI products. This can include termination of an account." No separate punishment specifically for a missing label is named; the consequence is the loss of monetization, since AI disclosure is part of the conditions of being monetizable There is no direct link "did not label — taken down/limited" on Snap's pages; there is the link "did not disclose — not monetized" source, checked 2026-08-04

Can you monetise it. Earning is possible on three conditions at once: "Content that leverages sophisticated editing or AI-based tools is monetizable—if it is original, entertaining, or informative, not misleading, and if the use of AI is disclosed anywhere in the content or creator profile." At the same time what is made entirely by AI outside Snapchat is shown worse: "Our content ranking algorithm rewards authentic, human-made content over wholly AI-generated content created outside of Snapchat, even when AI-generated content has transparency disclosures. AI-generated content that was created within Snapchat is eligible for recommendation." Not monetized is "content where the core visual or audio contribution is mechanically or automatically generated" An important detail: the label does not save one from being demoted in the feed — "even when AI-generated content has transparency disclosures". The rule about recommendations is on the page https://values.snap.com/policy/content-guidelines-recommendation-eligibility/recommendation-eligibility/quality. Monetization depends on being eligible for recommendation: "If you primarily or frequently publish content that is “not eligible for recommendation,” you are most likely not a good candidate for content monetization on Snapchat" source, checked 2026-08-04

Platform card

Threads

Is a label required. Meta's general rule applies: "We require people to disclose, using our AI-disclosure tool, whenever they post organic content with photorealistic video or realistic-sounding audio that was digitally created or altered, and we may apply penalties if they fail to do so." Threads is named directly among the apps where the label is applied: "in the coming months we will label images that users post to Facebook, Instagram and Threads when we can detect industry standard indicators that they are AI-generated" No separate Threads rule on AI was found: Threads has no "Community Standards" of its own, Meta's standards apply. Its own help article "Label AI content on Threads" (help.instagram.com/407718162047721) is unavailable: the domain returns HTTP 400 without a browser, and the robots.txt of help.instagram.com contains "User-agent: ClaudeBot / Disallow: /". Hence partial: the rule is confirmed at Meta level, but not on a page of Threads itself source, checked 2026-08-04

How it is labelled. The "AI info" label and a separate disclosure tool for the creator: "we’re adding a feature for people to disclose when they share AI-generated video or audio so we can add a label to it"; "Meta began adding labels to a wider range of video, audio and image content", "These are displayed as “AI Info”" The Transparency Center page speaks of "Facebook, Instagram, and Threads" together. Exactly what the toggle looks like in the Threads posting form has not been checked, the Threads help centre is closed (see the note to ai_disclosure_required) source, checked 2026-08-04

Does the platform detect it itself. Yes. "we’re building industry-leading tools that can identify invisible markers at scale – specifically, the “AI generated” information in the C2PA and IPTC technical standards"; "in the coming months we will label images that users post to Facebook, Instagram and Threads when we can detect industry standard indicators that they are AI-generated" Threads is named in the very title of the Meta blog entry. Both standards are named — C2PA and IPTC source, checked 2026-08-04

What is forbidden. There is no separate list of prohibitions for AI; Meta's standards apply: "We may place an informative label on the face of content – or reject content submitted as an advertisement – when the content is a photorealistic image or video, or realistic sounding audio, that was digitally created or altered and creates a particularly high risk of materially deceiving the public on a matter of public importance"; misinformation carrying a risk of violence, harmful medical misinformation and misinformation about elections and the census are removed regardless of how they were produced Threads has no list of AI prohibitions of its own; everything stated is Meta's rules, which extend to the app source, checked 2026-08-04

What happens if you break the rule. "…and we may apply penalties if they fail to do so" — Meta's general wording about punishment for a missing label; what exactly the punishment is has not been disclosed No penalties described specifically for Threads were found source, checked 2026-08-04

Platform card

TikTok

Is a label required. Required for realistic AI content. Word for word from the platform's blog: "The policy requires people to label AI-generated content that contains realistic images, audio or video, in order to help viewers contextualize the video and prevent the potential spread of misleading content"; "we label AIGC made with TikTok AI effects, and have required creators to label realistic AIGC for over a year" The primary source of the requirement is the community guidelines, the "Integrity and authenticity" section at https://www.tiktok.com/community-guidelines/en/integrity-authenticity. It cannot be read: the robots.txt of the www.tiktok.com domain contains the groups "User-agent: ClaudeBot", "Claude-User", "Claude-SearchBot", "anthropic-ai" with "Disallow: /". The help page support.tiktok.com/en/using-tiktok/creating-videos/ai-generated-content answers 302 to the same closed domain. The word-for-word wording of the rule is therefore taken from the platform's own blog source, checked 2026-08-04

How it is labelled. A separate labelling tool at publication, plus improvised ways are permitted: "we're rolling out a new tool for creators to easily inform their community when they post AI-generated content. The new label… they can apply it to any content that has been completely generated or significantly edited by AI"; "Creators can now do this through the new label (or other types of disclosures, like a sticker or caption)". On coverage: "We also built a first-of-its-kind tool to make this easy to do, which over 37 million creators have used since last fall" There is no step-by-step instruction ("where to click") in the platform's open sources: the help centre is closed by a redirect to www.tiktok.com, which is disallowed in robots.txt for Claude's agents. The wording "sticker or caption" — that is, a sticker or a caption — is named directly as an acceptable alternative to the tick box source, checked 2026-08-04

Does the platform detect it itself. Yes, in three ways. C2PA: "we’re expanding auto-labeling to AIGC created on some other platforms by launching the ability to read Content Credentials, a technology from the Coalition for Content Provenance and Authenticity (C2PA)… TikTok is the first video sharing platform to put Content Credentials into practice". Its own credentials: "we’ll also start attaching Content Credentials to TikTok content, which will remain on content when downloaded". Invisible watermarks: "we’ll start adding invisible watermarks to AI-generated content made with TikTok tools like AI Editor Pro, and content uploaded with C2PA Content Credentials"; "“Invisible watermarks” add another layer of safeguards with a robust technological “watermark” that only we can read, making it harder for others to remove". The label shows the origin: "we enhanced our AIGC labels by adding more context around whether content was labeled due to our AI detection, creator labels, or TikTok AI tools" On the invisible watermarks and the origin of the label — on the platform's second blog page: https://newsroom.tiktok.com/en-us/more-ways-to-spot-shape-and-understand-ai-content. There the platform names a figure: more than 3 billion videos labelled as AI content source, checked 2026-08-04

What is forbidden. "our policies firmly prohibit harmfully misleading AI-generated content—whether it's labeled or not"; "we… signed an industry pact this year to combat the use of deceptive AI in elections". Also: "We continue to refine our labeling approach as industry norms and expert guidance evolve, while maintaining firm policies against harmful AIGC" It was not possible to obtain the list of prohibitions (voices of real people, faces, elections, medical claims) in open form: it is in the community guidelines on www.tiktok.com, closed to Claude's agents in robots.txt ("Disallow: /"). The wording "whether it’s labeled or not" is important: the label does not lift the prohibition source, checked 2026-08-04

Platform card

Twitch

What is forbidden. Abuse of fakes is prohibited: "Creating “deepfake” videos to insult or abuse another person" (in the section about bullying, next to "Sharing negative doctored or artistic content to abuse or degrade another person"); "Creating or sharing doctored or ‘deepfake’ intimate images of an identifiable person" (in the section about sexual harassment). Nothing is said separately about the voices of real people, elections and medical claims in connection with AI; there is a general ban on "Impersonation of another person, company, or organization" and on "Harmful Misinformation Actors", not tied to the way of production The word "deepfake" is the only indication of AI in all the community guidelines. Both prohibitions are worded through harm to a particular person, not through the way of production source, checked 2026-08-04

Platform card

X (Twitter)

Is a label required. No general requirement to label all AI content was found. There is a requirement for one case — armed conflicts: "For any AI generated content depicting armed conflict, you must add a clear disclosure that it is AI generated" The page is closed by Cloudflare protection: both curl and WebFetch get HTTP 403 and the stub "Just a moment... Enable JavaScript and cookies to continue". The protection must not be bypassed. The words quoted are taken from the search index for this same official page, their exactness has not been re-checked against the page itself — hence confidence partial. The requirement relates to the "Content Monetization Standards", that is, to those who earn, and not to all creators source, checked 2026-08-04

How it is labelled. No separate labelling tool for the creator at publication is described: the rule requires that one "add a clear disclosure that it is AI generated", that is, disclosure inside the post itself. The label is applied by the platform: "For posts containing media determined to have been significantly and deceptively altered or fabricated, X may add a “Manipulated media” label" The page is closed by Cloudflare protection: both curl and WebFetch get HTTP 403 and the stub "Just a moment... Enable JavaScript and cookies to continue". The protection must not be bypassed. The words quoted are taken from the search index for this same official page, their exactness has not been re-checked against the page itself — hence confidence partial. Neither a tick box at upload nor a metadata field is described on the available X pages source, checked 2026-08-04

Does the platform detect it itself. Partly: "To determine if media have been significantly and deceptively altered or fabricated, X may use its own technology or receive reports through partnerships with third parties". No mentions of C2PA or IPTC were found The page is closed by Cloudflare protection: both curl and WebFetch get HTTP 403 and the stub "Just a moment... Enable JavaScript and cookies to continue". The protection must not be bypassed. The words quoted are taken from the search index for this same official page, their exactness has not been re-checked against the page itself — hence confidence partial. Nothing is said about content provenance standards (C2PA, IPTC) on the available X pages source, checked 2026-08-04

What is forbidden. "Media that is shared as authentic but is significantly and deceptively altered, manipulated, or fabricated in a way that fundamentally changes its meaning is prohibited, including media depicting a real person that has been fabricated or simulated, especially through use of algorithms or broader artificial intelligence". Not punished: "X will not take action to label or remove media that have been edited in ways that do not fundamentally alter their meaning, such as retouched photos or color-corrected videos" The page is closed by Cloudflare protection: both curl and WebFetch get HTTP 403 and the stub "Just a moment... Enable JavaScript and cookies to continue". The protection must not be bypassed. The words quoted are taken from the search index for this same official page, their exactness has not been re-checked against the page itself — hence confidence partial. The prohibition is worded through deception, not through the way of production: what matters is that the post is passed off as authentic ("shared as authentic") source, checked 2026-08-04

What happens if you break the rule. For a missing label on an AI video about an armed conflict: "Effective March 3, 2026 users who post AI-generated videos of an armed conflict without adding a disclosure that it was made with AI will be suspended from Creator Revenue Sharing for 90 days. Subsequent violations will result in a permanent suspension from the program". For misleading altered content the platform "may add a “Manipulated media” label", and also labels or takes down the post The page is closed by Cloudflare protection: both curl and WebFetch get HTTP 403 and the stub "Just a moment... Enable JavaScript and cookies to continue". The protection must not be bypassed. The words quoted are taken from the search index for this same official page, their exactness has not been re-checked against the page itself — hence confidence partial. This is the only numeric punishment for a missing AI label found at the platform: 90 days of exclusion from the revenue share, and permanently on a repeat source, checked 2026-08-04

Can you monetise it. "Monetized content, like other content on X, must be in compliance with X’s rules on Safety, Authenticity, and Privacy. You must only monetize content that is original and authentic, or otherwise content that you have the rights to monetize". Plus mandatory disclosure for AI content about armed conflicts (see ai_disclosure_required) The page is closed by Cloudflare protection: both curl and WebFetch get HTTP 403 and the stub "Just a moment... Enable JavaScript and cookies to continue". The protection must not be bypassed. The words quoted are taken from the search index for this same official page, their exactness has not been re-checked against the page itself — hence confidence partial. No direct ban on earning from AI content was found — originality, authenticity and disclosure in the stipulated case are required source, checked 2026-08-04

Platform card

YouTube

Is a label required. Required for photorealistic AI content. Word for word: "we require creators to disclose when they use AI to meaningfully alter or generate photorealistic content"; "Creators must disclose GenAI content that: Makes a real person appear to say or do something they didn’t do. Alters footage of a real event or place. Generates a realistic scene that didn’t actually occur." Not required: "Creators don’t need to disclose non-realistic content that’s made with AI, or edits to realistic content that are minor. Minor edits are ones that are primarily aesthetic, and don't alter the content in a way that could mislead the viewer about what actually happened." Examples without a label: "Applying beauty filters", "Color adjustment or lighting filters", "Production assistance, like using generative AI tools to create or improve a video outline, script, thumbnail, title, or infographic", "Caption creation", "Cloning one’s own voice to create voice overs or dubs". Examples with a label: "AI generated music", "AI generated realistic videos of a match between two real professional tennis players", "Making it appear as if someone gave advice that they did not actually give", "Depicting a public figure stealing something they did not steal" The page has been renamed: the title is now "Disclosing use of GenAI content" (previously "altered or synthetic content"). The lists of examples on the page are marked as incomplete ("this isn’t a complete list"). Obtained by an ordinary GET, the page opened source, checked 2026-08-04

How it is labelled. A tick box when uploading in YouTube Studio: "In the Attributes section, under “AI use,” select Yes if your content meets the disclosure requirements or No if not." What the label looks like to the viewer: "Label on the video player for photorealistic AI content"; "Labels may appear in the expanded description for AI content that is non-photorealistic or animated" The label is applied by the creator themselves at the "Attributes" step in Studio on a computer or in the mobile app. There is no separate help article about removing the label, apart from the "AI disclosure survey" source, checked 2026-08-04

Does the platform detect it itself. Yes. "YouTube may automatically apply an AI label on the video player or in the expanded description for: Content made using YouTube's GenAI tools; Content that contains C2PA metadata; Content that our internal systems detect is Al generated or altered." An automatic label cannot always be removed: "However, content made with YouTube’s AI tools, content containing C2PA metadata, or content labeled after manual review can not be adjusted." C2PA metadata and its own recognition systems are named directly. On an error in the automatic label: "If our systems make an error, creators will have the ability to change AI disclosure in most cases by selecting No in the AI disclosure survey" source, checked 2026-08-04

What is forbidden. The label does not cancel the prohibitions: "Creators should disclose when content is generated or meaningfully altered with AI, but this disclosure is not a free pass to impersonate a person, entity, or channel. Our Community Guidelines, including our impersonation policy, apply to all content regardless of how it is made." Prohibited: "It may also include using AI to copy the voice or likeness of an individual, to make it appear as if the channel is owned or authorized by that individual"; "Using someone's AI-generated likeness or voice to falsely imply they own or authorize your channel’s content"; "Using AI to make it look like a famous person is willfully participating in your video or endorsing a product"; "Misrepresenting an individual by using an AI version of their voice and false metadata to pass off a fake speech as real" This is the rule about impersonating someone else (the impersonation policy). The injured party can file a complaint: "If you find your personal information, a video featuring you, or AI-generated content mimicking your likeness or voice without your consent, you can file a request for removal under our privacy guidelines". There is no separate ban on AI content about elections on this page — it is in the rules on election misinformation source, checked 2026-08-04

What happens if you break the rule. "Creators who consistently choose not to disclose this information may be subject to manual application of a label, or penalties from YouTube, including removal of content or suspension from the YouTube Partner Program." At the same time the label itself is not punished: "Disclosing AI content won’t limit a video’s audience or impact its eligibility to earn money." The punishment is described for a systematic ("consistently") absence of the label, not for a single case. The page names no specific terms or thresholds source, checked 2026-08-04

Can you monetise it. You may earn on AI content if it is original and not template-based. Allowed: "Content that expresses your unique creative voice, like using AI to visualize a unique character and narrative you invented"; "Content that utilizes creative tools to assist in delivering a unique, well-researched, or creative narrative, like using AI to edit your video scripts or generate a unique background visual for your content". Not admitted to monetization: "AI-generated content made with generic or unoriginal templates giving the impression of mass production without adding the creator’s original, authentic insights or perspective"; "Content that lacks a clear narrative arc or logical progression, such as videos that stitch together unrelated or inconsistent AI clips to surprise or shock viewers"; "Content that uses deceptive or misleading imagery or narratives, such as realistic visuals tricking viewers into believing a fake celebrity death or natural disaster has occurred". A separate rule, "AI Personas Related to Sensitive Topics": "This policy refers to channels that use AI-generated personas to deliver information on sensitive topics… on topics such as health, legal issues, finances, or politics… channels uploading this content will not be allowed to monetize" — the examples: "An AI “doctor” providing medical diagnoses, health advice, or wellness remedies", "AI-generated podcast hosts offering financial guidance, investment tips, or wealth management advice", "AI personas giving legal advice or interpreting laws" On the same page in the change log: "July 15, 2025: We’re making a minor update to our “repetitious content” policy to better clarify this includes content that is repetitive or mass-produced. We are also renaming this policy from “repetitious content” to “inauthentic content.”" The phrase that the AI label does not affect income is at support.google.com/youtube/answer/14328491 source, checked 2026-08-04

Platform card

Dzen

What is forbidden. Automatically created content is itself prohibited, regardless of a label: Dzen prohibits content that is "wholly or almost wholly created automatically, without the participation of the author". The list includes "automatic translations from other languages without editing", "automatically generated texts", "automatically generated video voice-overs and audio tracks", and also material with "repeating words and phrases unconnected with each other". This is classed as spam — a "serious violation" This is the strictest framing of the question encountered: not "label it" but "it is not allowed". The Dzen help centre is assembled in the browser: an ordinary GET returns a shell of about 1 KB without text, so the pages were read through WebFetch. The robots.txt of dzen.ru does not disallow these addresses. Quoted text is our translation from Russian. source, checked 2026-08-04

What happens if you break the rule. For automatically created content: the posts are "shown only to subscribers"; further — "Dzen may block the post or even the whole channel — they will not be visible to anyone, even by a direct link"; "Channels and sites with a large amount of spam are unavailable on the platform" The punishment is imposed not for a missing label (none is even required) but for the very fact of automatic generation. The Dzen help centre is assembled in the browser: an ordinary GET returns a shell of about 1 KB without text, so the pages were read through WebFetch. The robots.txt of dzen.ru does not disallow these addresses. Quoted text is our translation from Russian. source, checked 2026-08-04

Can you monetise it. There is no separate rule about AI in the monetization terms. The general one: "If posts breach the requirements, Dzen may switch off their monetization. If there are many such breaches, the restriction extends to the whole channel". Since automatically created content is classed as spam and prohibited, it cannot be earned from The monetization rules page was read: it contains no words about neural networks, AI or automatically created content. The conclusion that earning is impossible follows from the ban in the rules about spam and not from a direct wording by the platform itself — hence partial Quoted text is our translation from Russian. source, checked 2026-08-04

Platform card

What to do

Label it if in doubt. Labelling is penalised nowhere, the absence of a label is penalised by many.
The absence of a rule is not permission. Правило может появиться завтра и подействовать на то, что published yesterday.
A platform label and an advertising disclosure are different things. The first says «a machine made this», the second «someone paid for this». The requirements for advertising disclosure are separate.
Check the monetisation rules, not only the publishing rules. Some platforms allow AI content to be published but not to be monetised.