What Meta Actually Requires for Creator Ads (And What the Internet Made Up)
One undated blog post is why everyone thinks Partnership Ads are mandatory. Here's what Meta's policy says, and what actually binds you.
For most of this summer, paid social teams have been told that Meta made Partnership Ads mandatory for every piece of creator content in paid, and that running UGC-style ads without the format is now a "Deceptive Practice" violation carrying an ad rejection and an account health penalty.
We went looking for the policy. It traces to a single undated, uncited paragraph on one compliance blog, which trade press then repeated. The same publisher's two other 2026 articles describe Partnership Ads as a consent-gated option, not a mandate. Meta has published no such rule that we can find.
That matters, because the obligations that are real and dated are getting less attention than the one that isn't.
TL;DR
- 📉 The "mandatory Partnership Ads" claim is single-sourced and undated. The publisher behind it states plainly that "Meta does not publish a single dated changelog of policy releases" and that its items are "observed enforcement themes rather than updates tied to specific effective dates."
- ✅ What Meta does require is narrower and verifiable: when promoting branded content integrations, advertisers must use the branded content tool to tag the business partner, per Meta's Advertising Standards.
- ⚖️ The FTC penalty everyone is quoting is out of date. The current maximum is $53,088 per violation, set by the FTC's 2025 inflation adjustment and unchanged for 2026 because OMB cancelled this year's adjustment.
- 🗓️ One real deadline passed 15 days ago. EU AI Act Article 50 transparency obligations became enforceable on 2 August 2026, with exposure up to €15 million or 3% of worldwide turnover.
- 💰 The commercial pull is real even if the mandate isn't. Meta's Partnership Ads hit a $10 billion revenue run rate in Q1 2026, more than doubling year over year.
- 🔑 The constraint nobody rumors about is supply. Partnership Ads permissions are a creator-consent problem, and consent is cheapest to secure at the brief, not at launch.
Is Meta requiring Partnership Ads for all creator content?
No verifiable Meta policy says that. Meta requires advertisers to use its branded content tool to tag a business partner when promoting branded content integrations, and it requires creator permission before a brand can run ads through a creator's handle. Those are real, documented requirements. A blanket mandate that every creator asset in paid must run as a Partnership Ad is not one we can source to Meta.
This distinction is not pedantic. A mandate would mean auditing and migrating every creator asset you have in market. The actual requirements mean something different and more manageable: get the tagging right, get permission properly, and review claims before you boost.
We have a commercial interest here. Cohley sells creator content operations, and a compliance panic would be good for us. We are writing the unexciting version anyway, because a paid social team that reorganizes its quarter around a rule that does not exist will not thank the vendor that sold them the panic.
Where the rumor came from
The chain is short. One compliance blog published a section headed "Mandatory Partnership Ads" stating that all influencer and creator content promoting a brand "must use Meta's Partnership Ads format." Trade coverage picked it up, attributed it to that analysis, and the claim entered circulation as settled fact.
Three things are worth knowing about the original.
It carries no effective date. Not in the section, not in the article's themes table, not in its FAQ. The same page opens by telling readers that Meta publishes no dated changelog and that everything below should be treated as observed enforcement patterns.
It is contradicted by its own publisher. That outlet's Instagram branded content guide, published in April 2026, treats organic branded content and Partnership Ads as two co-existing legitimate formats, and describes boosting as gated on the creator enabling permission, not on a format requirement.
The supporting details did not survive checking either. The widely repeated figure that undisclosed AI content accounts for 14% of Meta ad rejections appears nowhere in the source. That page states directly that "Meta does not publish a breakdown of rejection reasons by percentage" and asks readers to treat any ranking as directional. The related claim that Meta extended AI disclosure from political ads to commercial ads in March 2026 is also absent. The word "political" does not appear in the article at all.
Claimed versus verifiable
- Partnership Ads mandatory for all paid creator contentSingle-sourced, undatedMeta requires the branded content tool for tagging a business partner when promoting branded content integrations
- Non-compliant UGC ads = "Deceptive Practice" violationSingle-sourced, uncitedNo Meta policy text located using this classification for this scenario
- Immediate rejection plus account health penaltyAsserted, then undercutSame source states Meta "does not publish a numeric account health score with fixed thresholds"
- Undisclosed AI content = 14% of rejections, third-largest categoryNot foundSource explicitly denies that any percentage breakdown is published
- AI disclosure extended from political to commercial ads, March 2026Not foundSection is undated; "political" never appears in the cited article
- Brand is liable for creator claims in its adsVerifiableLong-standing FTC principle: the advertiser is responsible for claims it disseminates
- Platform label alone satisfies FTC disclosure—FTC expects disclosure inside the content, not only in platform metadata
What Meta's Advertising Standards actually say
Meta's Advertising Standards define branded content within ads as a creator or publisher's content that features or is influenced by a business partner for an exchange of value, and require that advertisers promoting branded content integrations use the branded content tool to tag the featured third party product, brand or business partner.
Read that carefully. It is a tagging and transparency requirement attached to promoting branded content. It is not a statement that every creator-produced asset must run under a creator handle.
Meta's actual direction of travel is visible in what it built rather than what it banned. On 11 December 2025 Meta expanded the Partnership Ads Hub to surface Instagram UGC and affiliate content from creators who tag or mention a brand, added a Facebook Partnership Ads API to identify suitable creator content at scale, expanded creator eligibility to Professional Mode profiles, and introduced a code-sharing flow so creators can grant ad permissions directly.
Meta is making the format easier to adopt, not compulsory. It has good reason to. Partnership Ads reached a $10 billion revenue run rate in Q1 2026, more than doubling year over year. Meta reports that adding partnership ads to a campaign delivers 19% lower CPAs and 13% higher click-through rates on average, and that 71% of consumers say they buy within a couple of days of seeing creator content across its apps. Those are Meta's own figures, and they should be read as such.
Independent analysis points the same way. A study of $130 million in ad spend found creator-led ads produced 19% more clicks, 10% higher conversions, and 5% lower costs than licensed UGC run from brand accounts. The category is growing underneath all of it: US creator ad spend reached $37 billion in 2025, up 26% year over year, according to the IAB.
You do not need a mandate to justify moving to Partnership Ads. The performance case already did that.
The permission requirement that actually bites
Here is the requirement that stops campaigns, and it is not the one making headlines: a brand cannot boost or promote a creator's content as a partnership ad unless that creator has granted ad permissions on that specific post.
Meta's December update made this easier by letting creators share an ad code with advertisers, including proactively and for content where the brand was never tagged. Easier is not the same as automatic. Someone still has to obtain the code, from the right creator, for the right asset, before the flight date.
This is the failure we see most often, and it is boring in a way that rumors never are. The asset is approved. The media plan is set. Then a permission is missing on four of the fifteen assets, and the launch slips while someone sends direct messages to creators who are asleep in another time zone.
Cohley collects TikTok Spark and Instagram Partnership Ad codes inside the creator upload flow, and boosting durations are set at the Brief stage rather than renegotiated at launch. That is a deliberately unglamorous design decision. It exists because chasing permissions after delivery is where paid social calendars go to die.
There is a second distinction that brands collapse constantly, and the cost of collapsing it rises the more creator content you run in paid: Meta ad permission and content usage rights are two different grants. Permission lets you run the ad on Meta. Usage rights govern where else the asset can live, for how long, in which territories, and in what derivative forms. A creator can grant one without the other. If your agreement covers only the first, your retail media placement, your product detail page, and your email header are all unlicensed uses.
Who's liable when a creator makes a claim in your ad?
You are. This is the part of the circulating story that holds up, though it predates any 2026 Meta policy and does not depend on one.
Under FTC rules, an advertiser is responsible for the claims it disseminates, regardless of who spoke them. If a creator says a serum cleared their acne in two weeks and you boost that content, the substantiation obligation is yours. The 2023 revised FTC Endorsement Guides remain the operative framework in 2026, and they are the most significant update to the agency's influencer disclosure standards since its first social media guidance in 2009.
For most creator programs, this is the genuine operational change, and it has nothing to do with ad formats. Content review has to include a claim check, not only a brand-fit check. Those are different skills, run by different people, against different criteria. A social lead can tell you whether an asset feels on-brand. Very few are briefed to read it the way a regulatory reviewer would.
The exposure is not theoretical. The FTC's rule on consumer reviews and testimonials, effective 21 October 2024, prohibits testimonials that misrepresent being from someone who does not exist or who never used the product, and the Commission has stated that advertising agencies, PR firms, and review brokers are not immune from liability under it.
One correction worth making, including to our own earlier writing. The penalty figure circulating in creator marketing content, including in our own comparison of AI-generated UGC and real creators, is $51,744. That figure is stale. The FTC's 2025 inflation adjustment raised the maximum to $53,088 per violation, effective on publication in the Federal Register on 17 January 2025. There was no 2026 increase: the government shutdown meant BLS never published October 2025 CPI-U data, so OMB cancelled the 2026 adjustment in Memorandum M-26-11. The correct current number is $53,088, and it is per violation, not per campaign.
Why the Paid Partnership label doesn't finish the job
Meta's Paid Partnership label satisfies Meta's platform-level disclosure requirement. It does not discharge your FTC obligation.
The FTC's position is that disclosure must be visible inside the content itself, not only in platform metadata that may not render in every viewing context. A viewer watching a Reel with sound off, in a feed, at speed, needs to understand the commercial relationship from the creative.
This creates a specific trap for anyone using the Partnership Ads Hub to boost existing organic posts. Content that cleared organic review months ago can still carry disclosure risk today, because the creator never built disclosure into the asset. Surfacing an old post inside the Hub does not retroactively make it compliant.
Practical guidance here is less settled than the confident checklists suggest. On the same platform, published three days apart, one compliance guide recommends verbal disclosure within the first 3 seconds of a Reel and another recommends within the first 10 seconds for video UGC ads. Both are framed as best practice rather than as an FTC rule, and the FTC-attributed version uses "or" between verbal and on-screen text, not "and."
The defensible reading: put the disclosure early, make it perceivable with sound off, and do not rely on the platform label alone. Then have counsel confirm the standard your category requires. Anyone quoting you an exact second count as a regulatory requirement is reading a blog, not a statute.
The obligations that are dated and enforceable
While the industry argued about a Meta rule nobody can source, three AI disclosure regimes came into force with actual dates attached.
EU AI Act Article 50 transparency obligations became enforceable on 2 August 2026. Not delayed, not deferred, as Goodwin put it plainly. Providers and deployers of systems that generate synthetic audio, image, video, or text must ensure outputs are marked in machine-readable format and detectable as AI-generated, and deployer-side deepfake disclosure duties took effect immediately. There is a transitional grace period for the marking obligation until December 2026 for generative systems placed on the market before 2 August. Exposure runs to €15 million or 3% of worldwide annual turnover. The European Commission's own guidance is the primary reference. The Act has extraterritorial reach, so a US brand whose AI-touched creative reaches EU audiences is in scope.
Two US regimes are also cited as in force, New York's synthetic performer disclosure law from June and a California metadata mandate from August. We have seen both referenced repeatedly in vendor content, and we have not yet verified either against statutory text. We are not going to restate their penalty figures here on the strength of secondary sources, which is exactly the mistake this article is about.
If your creative uses AI to generate or substantially modify a human likeness, voice, or performance, you have a dated obligation. If AI is doing color grading, transcription, or workflow, you generally do not. The consumer evidence supports keeping that line clean regardless of law: in a peer-reviewed 2026 experiment, identical review text scored 4.18 out of 5 on consumer trust with no AI label, 3.56 labeled "AI-assisted," and 2.30 labeled "AI-generated". Only 35% of US consumers say they trust AI-generated content and 91% expect brands to disclose AI use, while 31% distrust AI-generated content entirely.
What this should change in your brief
Every requirement in this article is cheaper to satisfy at the brief than at the ad account. That is the whole practical argument.
Where each requirement is cheapest to handle
- Meta ad permissionAd code captured at upload; boosting duration set upfrontDirect-message chase across time zones during launch week
- Usage rights scopeChannels, duration, territory, derivatives defined before the shootAsset cannot legally run on the placement you bought
- Claim substantiationClaim limits written as observable non-negotiables in the briefLegal review after delivery, then a reshoot
- In-content disclosureDisclosure specified as a deliverable requirementAsset unusable in paid, or usable with risk
- AI use in productionCreator declares AI tools usedYou inherit an undisclosed obligation you cannot see
Two structural notes on making this real.
Requirements only work when they are observable. "On-brand" is not a requirement. "Verbal disclosure in the opening line, spoken, before any product claim" is. We wrote a longer piece on how to write creator content requirements that actually work, because vague requirements are the single largest source of change requests we see.
And the supply problem is the one that will actually cost you a quarter. If you do need to pause or replace creative, the constraint is not the ads manager. It is whether you have enough approved, rights-cleared assets to keep spending while you fix things. Rhone built a library of 737 assets and moved from 10 to 15 creative tests per week to 50 per day, with a 3 to 4 times increase in ROAS against internal content. That kind of depth is what makes a compliance change survivable rather than disruptive. Cohley is a badged Meta Business Partner, and the honest version of what that means is access and integration, not a compliance guarantee. Our AI Asset Analysis checks submitted assets against the requirements you set and flags them for a human reviewer. It checks requirements. It does not verify legal or FTC compliance, and no vendor's automated check does.
Who this is for
If you run creator content in Meta paid and you were about to pause campaigns: don't, on the strength of the mandate story. Confirm the tagging and permission requirements on your active assets, then move on to claim review, which is the real gap.
If you are in a regulated category such as supplements, beauty, health, or financial services: the claim liability section is your priority. Add a substantiation checkpoint between creative approval and boost. This is worth doing this month regardless of what Meta does next.
If you sell into the EU: Article 50 is live now and is the only firmly dated obligation in this article. Audit any creative where AI generated or modified a person.
If you are an agency running this across a client portfolio: every client account carries its own exposure and its own creator agreements. Our Power Digital case study covers what centralizing creator operations across hundreds of client accounts looked like: 5 times faster sourcing, roughly 75% shorter production timelines.
If your creator agreements predate 2026: check whether they say anything about AI modification of delivered footage. Most do not, and running human creator footage through a tool that regenerates voice or likeness can move a compliance-safe asset into scope.
FAQ
Are Meta Partnership Ads mandatory in 2026?
No verifiable Meta policy establishes a blanket mandate. Meta requires advertisers to use the branded content tool to tag a business partner when promoting branded content integrations, and requires creator permission to run ads through a creator's handle. The widely repeated "mandatory" claim traces to one undated, uncited secondary source.
Do I need Partnership Ads for gifted product content?
Meta's tagging requirement applies to branded content integrations, which are defined by an exchange of value, and gifted product is an exchange of value. Separately, gifted product triggers FTC disclosure obligations for the creator. Treat gifting as a material connection requiring disclosure regardless of format.
Does Meta's Paid Partnership label satisfy FTC disclosure requirements?
No. The label satisfies Meta's platform-level requirement only. The FTC expects disclosure to be visible within the content itself, not solely in platform metadata that may not appear in every viewing context. Build disclosure into the creative and treat the label as supplementary.
Who is liable if a creator makes a false product claim in my ad?
The advertiser. Under FTC rules, a brand is responsible for the claims it disseminates regardless of who said them. Boosting a creator's claim makes the substantiation obligation yours, which means content review needs a claim check separate from a brand-fit check.
What is the current FTC penalty per violation?
$53,088, set by the FTC's 2025 inflation adjustment effective 17 January 2025. There was no 2026 increase because OMB cancelled the adjustment after October 2025 CPI data was not published. Penalties apply per violation, not per campaign. Older content citing $51,744 is out of date.
Does the EU AI Act apply to a US brand?
Yes, if AI-generated or AI-modified output reaches EU audiences. Article 50 has extraterritorial scope and became enforceable on 2 August 2026, with penalties up to €15 million or 3% of worldwide annual turnover. A machine-readable marking grace period runs to December 2026 for systems placed on the market before that date.
If I use AI to edit a real creator's video, do I have to disclose it?
It depends on whether the AI alters the person. Color grading, noise removal, captions, and background work generally do not trigger disclosure. Generating speech the creator never recorded, cloning their voice, or substantially reshaping their likeness does. Ask creators to declare AI tools used at delivery.
Next step
Pull your active creator assets in Meta paid and sort them into three piles: assets with confirmed ad permission, assets where permission is assumed but undocumented, and assets carrying a product claim nobody has substantiated. The second and third piles are your actual exposure, and neither one requires a Meta mandate to be a problem.
If the third pile is large and you have nothing cleared to replace it with, that is a supply problem, and it is the one we solve. Book a demo and we will walk through what capturing permissions and claim requirements at the brief stage looks like on your programs.