Temu has until August 28, 2026 to hand the European Commission a plan explaining how it will fix the risk-assessment failures that earned it a EUR 200 million fine under the Digital Services Act. The deadline is the pivot point between a penalty already imposed and a supervision regime that can bill the marketplace daily until Brussels is satisfied.
In short
- The deadline: Temu must submit a remediation action plan by August 28, 2026 under the DSA’s enhanced supervision procedure.
- The penalty already levied: a EUR 200 million fine announced in May 2026 for breaching the systemic risk-assessment duties in Articles 34 and 35.
- What follows: the European Board for Digital Services gets one month to opine, then the Commission gets one further month to decide and set an implementation period.
- The downside risk: an inadequate plan exposes Temu to periodic penalty payments of up to 5% of average daily worldwide turnover, a meter that runs rather than a one-off charge.
- The pattern: Temu’s case sits inside a sequence that now includes a EUR 550 million AliExpress fine and open proceedings against Shein, all built on the same theory: the paperwork is the offense.
What Temu has to file by August 28
The document due in Brussels is not an appeal and not a settlement. It is a remediation action plan: a description of the specific measures Temu intends to take to bring its systemic risk assessment into compliance, plus the timetable for each. The Commission set the date when it adopted the non-compliance decision in May 2026, and the clock has run since.
The distinction matters because the fine and the plan are separate instruments. Paying EUR 200 million does not close the file. Under the DSA, a non-compliance decision automatically opens a supervision phase in which the platform has to demonstrate, on a schedule the regulator controls, that the underlying defect has been repaired.
Temu has said it disagrees with the decision. In a statement reported at the time of the fine, the company called the penalty disproportionate and said the decision “relates to our first DSA assessment in 2024 and does not reflect the current state of our systems.” That position, if maintained, complicates the drafting: a plan has to concede enough to satisfy the Commission without undermining any challenge to the underlying finding.
The Commission did not order Temu to remove a defined set of listings. It found that the platform’s assessment of its own systemic risks was methodologically inadequate, which is a governance finding rather than a content finding. The remedy therefore has to be a governance artifact: a rebuilt assessment methodology, evidenced with platform-specific data.
That is a harder deliverable than a takedown sweep. Removing listings is measurable and fast, and a platform can show the numbers within days. Proving that a risk-assessment process is now diligent requires the Commission to accept a methodology, and methodologies are argued over rather than counted.
How the Commission built the EUR 200 million case
The decision rests on a body of evidence the Commission assembled itself rather than on complaints forwarded from national authorities. That is a notable shift in how these cases are made, and it explains why the finding was framed around assessment quality. A regulator that has run its own tests can compare what it found against what the platform said it expected to find.
The mystery shopping evidence
Commission investigators ran independent purchases of products listed on the platform and put them through safety testing. According to reporting on the decision, a very high percentage of the chargers bought during the exercise failed basic electrical safety tests. Baby toys performed similarly badly, with a high share presenting risks rated between medium and high severity.
The specific defects cited were chemical content exceeding legal safety limits and detachable parts small enough to present a suffocation hazard. Neither category is exotic: chargers and children’s products are the two most heavily regulated consumer goods classes in the single market. The choice of test categories was therefore not accidental.
Buying the goods rather than scraping the listings also closes an evidentiary gap that platforms have historically exploited. A listing that describes a compliant product proves nothing about the item that arrives. Physical testing removes the argument that the catalog data looked fine.
Where the 2024 risk assessment fell short
The core finding is that Temu’s first DSA risk assessment leaned on general information about the e-commerce sector rather than evidence drawn from Temu’s own operations. The Commission concluded that this approach seriously underestimated the exposure of European consumers to illegal products on the platform.
The gap between a sector-level assessment and a platform-level one is the whole case. A marketplace that sources predominantly from a long tail of third-party sellers in a single origin market does not carry sector-average risk. Assessing it as if it did produces a number that is wrong in a predictable direction.
What a compliant assessment looks like follows from that critique. It would draw on the platform’s own enforcement telemetry, category-level defect rates observed in its own returns and complaints data, seller-cohort risk scoring, and testing of its own goods rather than published sector studies. Each of those inputs is something the platform already holds.
This is the same analytical move that has shaped the broader regulatory push. Our earlier coverage of how marketplace safety enforcement runs through customs data traced the same logic from a different direction: regulators increasingly want platform-specific evidence, not category averages.
Recommender systems and influencer promotion
The third strand of the finding concerns amplification. The Commission held that Temu’s assessment did not properly examine how the design of its own service could spread illegal products faster than a neutral catalog would. Two mechanisms were named: the recommender systems that decide what a shopper sees, and the influencer-driven promotion programs that push specific items.
Treating affiliate and influencer marketing as a product-safety risk factor rather than purely a marketing channel is the genuinely new element here. It converts a growth function into a compliance surface. Any marketplace running an affiliate program in the EU now has a documented precedent that the program belongs inside the risk assessment.
The engineering consequence is that risk assessment stops being an annual report written after the fact and becomes an input to product decisions. If ranking weights and affiliate payouts influence the distribution of non-compliant goods, then changes to either need a compliance review before they ship. Several large marketplaces have already restructured toward that model.
What Article 75 enhanced supervision actually does
The DSA’s enforcement toolkit sits in Articles 64 to 76, and the enhanced supervision procedure is the mechanism that turns a one-off fine into an ongoing obligation. It applies specifically after a non-compliance decision against a very large online platform. The general obligations themselves entered into force on February 17, 2024, so the regime is now past its first full assessment cycle.
The two-month clock after submission
Once Temu files, the European Board for Digital Services has one month to issue an opinion on whether the proposed measures are adequate. The Commission then has one further month to adopt a final decision and set the period within which the measures must be implemented. On the stated timetable, that puts a Commission verdict in the window running from late September into late October 2026.
The Board’s opinion is not binding, but it is the point at which national digital services coordinators get a formal say. Several of those authorities have their own consumer-protection dockets involving the same platforms, which raises the chance of an opinion that pushes for more rather than less.
It is also the stage where the file becomes partially visible. Commission decisions are published in some form, and the implementation period it sets will be a public number. That number is what the rest of the sector will read as guidance.
What periodic penalty payments would cost
If the Commission concludes the plan is inadequate or that Temu has not implemented it, the escalation is not another fixed fine. It is periodic penalty payments, which under the DSA can reach 5% of average daily worldwide turnover for each day of delay. The ceiling on the headline fine is separate: up to 6% of total worldwide annual turnover for a breach.
The structural difference matters more than either percentage. A fine is a number a large platform can provision for and appeal. A daily accrual creates pressure to settle the substance quickly, because the cost of arguing rises every day the argument continues.
That asymmetry is deliberate. The DSA’s drafters were working against a decade of competition cases in which large fines were absorbed, litigated for years, and produced limited behavioral change. Tying the meter to daily turnover was the answer to that history.
How Temu’s case compares with AliExpress and Shein
Temu’s file is one of three running against China-linked marketplaces, and the three are at visibly different stages. Reading them together gives a clearer picture of the enforcement arc than any one case does alone.
| Platform | Status | Penalty | Key date | Focus of the case |
|---|---|---|---|---|
| Temu | Non-compliance decision, enhanced supervision | EUR 200 million (May 2026) | Action plan due August 28, 2026 | Systemic risk assessment quality (Articles 34 and 35) |
| AliExpress | Non-compliance decision, enhanced supervision | EUR 550 million (July 20, 2026) | Action plan due October 20, 2026 | Detection and enforcement systems for illegal goods |
| Shein | Formal proceedings open, no decision | None to date | No public deadline | Illegal products, addictive design, recommender transparency |
| X | Non-compliance decision (non-marketplace comparator) | EUR 120 million (December 2025) | Concluded phase | Platform transparency obligations |
The AliExpress case, which drew a EUR 550 million penalty and displaced Temu’s fine as the largest issued under the regime, was framed slightly differently. The Commission there found that counterfeit goods, unsafe toys and dangerous cosmetics stayed live on the platform for weeks because detection and enforcement systems were inadequate. Our report on the record EUR 550 million DSA penalty against AliExpress set out how that decision was structured.
The two theories are complementary rather than alternative. Temu was faulted for not seeing the risk clearly; AliExpress was faulted for not acting on what it saw. A platform can fail either test independently, which means satisfying one does not insulate against the other.
Shein remains the open variable. Formal proceedings opened in February 2026 covering addictive design, recommender system transparency and the sale of illegal products, and no fine has followed. That the Commission has taken longer over Shein than over Temu suggests either a broader theory of the case or a more contested evidentiary record.
Three marketplace files in roughly fifteen months, two of them now at the remedy stage, is a cadence rather than a coincidence. For any marketplace approaching the 45 million monthly active user threshold that triggers very large online platform designation, both theories are now live risks. The designation itself is mechanical: cross the threshold, report it, and the obligations attach.
Why the risk assessment, not the products, is the battleground
It would be easy to read these cases as being about unsafe chargers and toys. They are not, or at least not primarily. The legal finding against Temu concerns whether the company looked properly, not solely whether it found everything.
That framing has a practical consequence for platforms: you cannot fully discharge the obligation by improving takedown speed. A platform with excellent removal metrics and a generic, sector-averaged risk assessment is still exposed under the theory the Commission has now tested successfully.
It also changes what evidence looks like in these files. The Commission’s mystery shopping results function less as a list of violations to be remedied and more as proof that the platform’s own risk estimate was wrong. The products are the control experiment, not the charge sheet.
For compliance teams, the reframing is uncomfortable because it removes a familiar defense. Demonstrating good faith effort has historically been persuasive with regulators. Under this reading, effort applied through a flawed methodology does not cure the flaw.
What changes for marketplace sellers
Sellers are not parties to the Commission’s proceedings, but they absorb the consequences of the remedies. When a platform tightens its risk controls under regulatory supervision, the tightening lands on listings, categories and onboarding requirements.
| Area | Typical pre-enforcement practice | What supervision pushes toward |
|---|---|---|
| Seller onboarding | Self-declared trader details, light verification | Documented traceability checks before listing goes live |
| Regulated categories | Listing first, compliance documents on request | Test reports and conformity documents required up front |
| Repeat offenders | Account-level suspension, easy re-registration | Systems designed to block rogue trader reappearance |
| Affiliate and influencer promotion | Growth channel, separate from compliance review | Inside the risk assessment, with category restrictions |
| Recommender placement | Optimized for conversion | Constrained for higher-risk product classes |
The categories most affected are the ones the Commission tested. Chargers, power adapters, batteries and other electrical accessories sit in the highest-scrutiny tier, alongside toys and anything intended for children under three. Sellers in those categories should expect document requests rather than warnings.
Very few compliant sellers will lose access outright. The realistic effect is added cost per listing: conformity documentation, testing where none was previously held, and slower time-to-live for new products. For thin-margin sellers moving high volumes of low-value goods, that cost is not trivial.
There is a competitive wrinkle. Sellers who already hold conformity documentation, typically those supplying EU retail through conventional channels, gain relative advantage when documentation becomes a listing prerequisite. Compliance cost is regressive across seller size, and consolidation toward larger sellers is the predictable outcome.
This compounds a customs squeeze already in progress. The parcel economics that made the model work have been reset from the trade side as well, a shift we covered when tracing how the bill for Temu and Shein finally landed in 2026.
How this stacks with the customs and tariff squeeze
The DSA file is arriving on top of, not instead of, the trade-side pressure. Low-value import treatment has tightened on both sides of the Atlantic, and the cost per parcel has moved accordingly. A platform absorbing both a duty change and a compliance rebuild is managing two independent cost lines at once.
The strategic response visible across the sector has been localization: EU warehousing, local seller recruitment, and shipment consolidation that moves goods across the border in bulk rather than parcel by parcel. That shift also happens to reduce exposure to the customs-linked enforcement channel, which is a secondary benefit rather than the primary driver.
We set out the logic of that pivot in detail when arguing that Temu and Shein would move to EU local fulfillment. The DSA remedy phase gives that argument an additional push, because a locally sourced catalog is easier to evidence in a risk assessment than a cross-border long tail.
Customs enforcement and platform regulation are converging on the same underlying requirement: know who your sellers are and what they are shipping. Trader traceability under the DSA and importer-of-record accuracy under customs law demand overlapping data. Platforms that build one system to satisfy both will spend less than those running parallel programs.
What the numbers say about Temu’s EU exposure
Scale is what brought Temu inside the regime and what determines the size of any penalty. The platform was designated a very large online platform on May 31, 2024, having reported user numbers above the 45 million monthly active user threshold.
| Measure | Figure | Source basis |
|---|---|---|
| VLOP designation date | May 31, 2024 | Commission designation decision |
| Designation threshold | 45 million monthly active users in the EU | DSA statutory threshold |
| Reported EU MAU, September 2024 | 92 million | Figures reported to the Commission |
| Reported EU MAU, December 2024 | 93.7 million average | Platform transparency reporting |
| Maximum DSA fine | 6% of total worldwide annual turnover | DSA enforcement framework |
| Maximum periodic penalty | 5% of average daily worldwide turnover | DSA enforcement framework |
Industry trackers put Temu’s current EU monthly user base considerably higher than the 2024 reported figures, though those estimates are not regulatory filings and should be treated as indicative. What is not disputed is the direction: the user base has grown since designation, which raises rather than lowers the relevance of the obligations.
For context on aggregate reach, reporting around the AliExpress decision noted that roughly one in five Europeans shops on Temu, Shein or AliExpress at least once a month, and put AliExpress at 193 million European users over the prior year. Whatever the precise figures, the three platforms together represent a consumer-facing surface large enough to explain the Commission’s prioritization.
The turnover base matters as much as the user base, because both penalty ceilings are expressed against worldwide turnover rather than EU revenue. For a platform whose parent operates a much larger domestic business, the theoretical maximum is calculated on a figure far exceeding anything earned in Europe. That asymmetry is a design feature of the regime, not an oversight.
The open strands that did not close in May
The May decision resolved one question out of several. When the Commission opened formal proceedings against Temu in October 2024, the scope was broader than risk assessment alone, and the remaining strands are still live.
Those strands cover the systems meant to prevent rogue traders from reappearing after suspension, addictive design features including game-like reward mechanics, recommender system transparency under Articles 27 and 38, and researcher access to platform data under Article 40. Each could generate a separate finding.
That matters for how the August 28 filing is read. A plan addressing only the risk-assessment defect leaves the other strands untouched, and the Commission retains the option to act on them independently of the enhanced supervision process. A platform in that position has to decide whether to remediate narrowly or pre-empt the open questions.
Temu’s public position also leaves open a challenge to the decision before the EU courts. Filing an action plan does not waive that right, and platforms in comparable positions have run both tracks simultaneously. An annulment action would take years, well beyond the supervision timetable, so the remedy is likely to be implemented regardless of whether the fine is contested.
What to watch between August 28 and year-end
The next four months contain a defined sequence of checkpoints, which is unusual in platform regulation and makes the file relatively easy to track.
- August 28, 2026: Temu’s action plan is due. Watch whether the company publishes any part of it, and whether the framing concedes the methodology point.
- Late September 2026: the European Board for Digital Services opinion window closes.
- Late October 2026: the Commission’s final decision window closes, including the implementation period it sets.
- October 20, 2026: AliExpress files its own action plan, giving a direct comparison of how two platforms answer similar findings.
- Open: any Commission move on the remaining Temu strands, or a first decision in the Shein proceedings.
The AliExpress overlap is the most informative item on that list. Two remediation plans filed eight weeks apart, against decisions built on related but distinct theories, will show whether the Commission is converging on a standard template for marketplace remedies or handling each case on its own terms.
For sellers and competing platforms, the useful signal is not the fine but the implementation period the Commission sets in late October. That figure will indicate how quickly Brussels expects a very large marketplace to rebuild a risk-assessment function, and it will become the informal benchmark everyone else is measured against.
The financing pressure that this regulatory sequence has already put on the sector is visible elsewhere too, including in how the EU crackdown has weighed on Shein’s IPO valuation. Enforcement risk that once sat in a footnote now shows up in the price investors are willing to pay.
Frequently asked questions
What exactly is due on August 28, 2026?
A remediation action plan submitted to the European Commission, setting out the measures Temu will take to correct the systemic risk-assessment failures identified in the May 2026 non-compliance decision, together with a timetable for implementing them.
Does the EUR 200 million fine go away if the plan is accepted?
No. The fine and the action plan are separate instruments. The penalty stands regardless of the plan, unless it is overturned by the EU courts. The plan governs what happens next in the supervision phase.
What happens if Temu misses the deadline or files an inadequate plan?
The Commission can impose periodic penalty payments of up to 5% of average daily worldwide turnover for each day of delay in complying. Unlike a fixed fine, that accrues continuously until compliance is achieved.
Which DSA provisions did the Commission find were breached?
The decision centers on Articles 34 and 35, the systemic risk assessment and risk mitigation obligations that apply to very large online platforms, with the specific finding resting on Article 34(1)(a).
How does this compare with the AliExpress fine?
AliExpress was fined EUR 550 million in July 2026, a larger penalty built on a finding that its detection and enforcement systems left illegal goods live for weeks. Its action plan is due October 20, 2026. Temu’s case turns on the quality of its risk assessment rather than removal performance.
Is Shein facing the same process?
Not yet at the same stage. The Commission opened formal proceedings against Shein in February 2026 covering illegal products, addictive design and recommender transparency, but no non-compliance decision or fine has been issued, so no action plan deadline applies.
What does this mean for sellers on the platform?
Expect stricter onboarding verification, up-front conformity documentation in regulated categories such as electrical accessories and children’s products, and tighter controls on affiliate promotion. The main near-term effect is added cost and slower listing times rather than loss of access.
Can Temu appeal the decision?
Yes. The company can bring an annulment action before the EU courts, and filing an action plan does not waive that right. Court proceedings would run on a multi-year timetable, well past the supervision milestones scheduled for autumn 2026.
Why does the Commission care about recommender systems in a product-safety case?
Because the decision treats amplification as part of the risk. If ranking algorithms and influencer promotion programs increase the reach of illegal listings, the Commission’s position is that their effect belongs inside the platform’s systemic risk assessment.