Temu may have spent as much as $962 million on Meta partnership ads shown through creator accounts in the UK and European Union, according to research by the Czech non-profit Online Risk Labs. The headline is dramatic, but it needs two important qualifiers: the spending figure is an estimate derived from ad reach rather than Temu invoices, and the researchers found signals of likely inauthenticity rather than proof that Temu created or knowingly controlled fake people.
Key takeaways
- Online Risk Labs estimated that Temu spent up to $962 million on Meta partnership ads across the UK and 27 EU countries from January 2025 to April 2026.
- The top 100 creator accounts handled 74% of Temu’s partnership advertising during the final four months studied; only eight were identity-verified.
- Frequent handle changes, unusual account locations and industrial-scale ad output led researchers to describe many as possible “burner creators”.
- The study does not establish that Temu created the accounts, knew they were inauthentic or ran fraudulent ads.
Everyone else is reporting a near-billion-dollar ad bill; we are explaining the mechanism that made a small set of creator identities function like an industrial advertising layer. That distinction matters because Meta partnership ads can make a brand-funded message appear under a creator’s identity. The commercial relationship is disclosed in the ad format, yet the account still supplies the human face and social proof that a normal corporate page cannot.
What the Temu research actually found
Fortune reported that Online Risk Labs analysed advertising activity from 31 official Temu pages across the UK and all 27 EU member states. The data covered January 2025 through April 2026 and included more than nine million ads with a cumulative reach of about 134 billion. Cumulative reach is not the same as unique people: one user can be counted many times after seeing many ads.
The researchers estimated spending from reach and average advertising costs because Meta’s public data does not reveal a single audited invoice total. That produced the upper estimate of $962 million. Two advertising specialists who reviewed the method told Fortune that the estimate was plausible, but Temu did not confirm it. A responsible reading therefore treats $962 million as a modelled ceiling, not booked expenditure.
The strongest finding is concentration. From January through April 2026, just 100 creator accounts accounted for 74% of Temu partnership ads in the region. They ran more than 1.4 million individual ads and generated 16.9 billion in cumulative reach, according to the research. Seventy-three of those accounts had changed their handles at least once, one changed its name 15 times, and only eight had an identity that researchers could verify.
How Temu partnership ads turn identity into distribution
A Meta partnership ad begins with content associated with a creator, publisher or other partner. A brand can then pay to distribute that content more widely through Meta’s advertising system. Meta’s own partnership-ad training presents the format as a way to combine creator credibility with paid reach.
The mechanism is attractive because it separates three jobs. The creator account supplies a recognisable identity and piece of content. The advertiser supplies money, targeting and performance goals. Meta supplies the auction, delivery system and measurement. When all three are genuine and disclosed, the arrangement is ordinary creator marketing. The risk appears when the identity layer is disposable, synthetic or difficult to verify.
Online Risk Labs found that 28 of the top 100 accounts were listed as based in Russia and 19 in China, although the campaigns targeted European audiences. Another 34 were US-based, while only six were based in the EU or UK. Location alone proves neither deception nor state involvement; remote marketing teams and global agencies are common. Combined with repeated handle changes, copied formats and extraordinary output, however, it becomes a reason for scrutiny.
Why the $962 million number is easy to misunderstand
The current draft originally implied that Temu paid fake creators nearly $1 billion. The evidence supports a narrower claim: researchers estimated total Meta partnership-ad spending tied to the studied Temu pages, then examined the creator accounts carrying a large part of that advertising. The study does not publish a ledger assigning the entire amount to fake identities.
| Claim | Supported by available evidence? | What readers should understand |
|---|---|---|
| Temu partnership ads cost up to $962m | Estimated, not confirmed | Derived from reach and average ad costs |
| 73 of the top 100 accounts showed warning signs | Yes, according to ORL | Handle changes are a signal, not final proof |
| Temu created fake creators | No evidence published | Control and knowledge remain unresolved |
| The ads were fraudulent | No | Fortune explicitly said there was no evidence of ad fraud |
| EU regulators received the findings | Yes, according to ORL | A submission requests assessment; it is not a ruling |
This distinction is especially important in technology reporting, where a striking total can outrun the method behind it. Our earlier explanation of the ChatGPT ads run rate made a similar separation between an annualised pace and realised yearly revenue. Numbers need the period, unit and calculation attached.
What Meta, Temu and European regulators have said
Fortune reported that Temu did not respond to its requests and that Meta declined to comment on the ORL findings. That silence does not verify the researchers’ interpretation. It leaves key questions open: who operated the creator accounts, whether agencies or affiliates were involved, how Meta verified partnership eligibility and what Temu knew about account ownership.
ORL said it submitted its material to the EU authority responsible for Digital Services Act oversight. The DSA requires large platforms to assess systemic risks and maintain ad transparency. The European Commission explains that online advertisements must be clearly labelled and identify who placed them and why a user saw them.
Temu already sits within a wider DSA enforcement context. The Commission opened formal proceedings in 2024 concerning illegal products, potentially addictive design, recommendation systems and researcher access. In May 2026 it imposed a €200 million penalty over Temu’s risk assessment for illegal products. That case is separate from ORL’s creator-account allegations and should not be presented as a ruling on them.
Lapaas Voice’s report on new Digital Services Act designations explains why scale changes compliance duties. A platform or marketplace does not merely react to individual violations; it must assess whether its systems create repeatable risks.
Why this matters beyond Temu
The episode exposes an accountability gap in performance advertising. Platforms can verify that a business paid for an ad without establishing that the public-facing creator identity represents a stable person. Brands can optimise campaigns through agencies and automated tools while remaining several contractual steps away from the account that users see. Creators can be real, synthetic, shared, renamed or abandoned.
The result is not necessarily illegal, but it can weaken the meaning of “creator”. If one account can push tens of thousands of campaign variations, change its name repeatedly and operate primarily as an ad vessel, audiences are not encountering the same relationship they expect from an independent reviewer. The account has become infrastructure.
Temu’s estimated $962 million partnership-ad push is best understood as a test of identity accountability: when paid distribution scales faster than platforms verify the people supplying social proof, a creator format can become an industrial advertising channel.
That is also why the story differs from ordinary influencer misconduct. A single hidden sponsorship is a disclosure problem. A network of high-output, unstable identities is a systems problem involving eligibility rules, agency oversight, ad libraries and regulator access to data.
What evidence would settle the dispute
The next useful step is not another estimate. Regulators and independent researchers need account-level records showing ownership, verification, agency relationships, payment flows and changes of control. Meta can explain how an account becomes eligible for partnership ads and whether repeated handle changes trigger review. Temu can disclose how it selected partners and what warranties its agencies or affiliates provided.
ORL should also publish enough methodology for others to reproduce the spend range and account classifications without exposing personal data. Reproducibility would help separate strong indicators from subjective judgement. If independent teams reach similar results from the same ad-library records, confidence rises.
Until then, readers should avoid two opposite errors. It is wrong to state that Temu secretly paid $962 million to invented people. It is also wrong to dismiss the concentration and identity signals as meaningless. The defensible conclusion is that a large, highly concentrated partnership-ad system deserves verification because the human layer may not be as human as it appears.
FAQs
Did Temu spend $962 million on fake creators?
No confirmed invoice or payment ledger proves that wording. Online Risk Labs estimated up to $962 million in Temu partnership-ad spending across the UK and EU, then found likely inauthenticity signals among many of the top creator accounts carrying those ads.
What is a Meta partnership ad?
It is paid advertising that uses content and identity associated with a creator or other partner while a brand funds distribution through Meta’s ad system.
How many Temu creator accounts were verified?
Online Risk Labs said only eight of the top 100 accounts in its analysis had identities it could verify. That does not mean all remaining 92 were fake; it means verification was missing or warning signals existed.
Is the European Union investigating this specific claim?
ORL says it submitted findings to the relevant DSA authority and requested a risk assessment. That is not the same as a formal finding or penalty connected to the creator-account research.
Get the day’s top stories in your inbox
One concise email. No spam, unsubscribe anytime.
