The employment presumption is not a dogma: what the Extremadura ruling on Glovo and the Flex model teaches
The Extremadura High Court ruling on Glovo's Flex model teaches how the employment presumption for riders is rebutted.

The Extremadura High Court judgment declaring the inexistence of an employment relationship between Glovo and its riders is more important for what it teaches about the mechanism than for the outcome of the case. The employment presumption for digital delivery platforms was born to correct an evidentiary imbalance: the company designs the model and keeps the data, so it must be the one to prove the rider's autonomy. The Extremadura ruling shows that the presumption works exactly like that: it is not an anticipated conviction, but a burden of proof that can be overcome.
The presumption is rebutted with design, not with words
The decisive detail of the judgment, according to the Poder Judicial note, is the list of elements the Chamber considered proven: freedom to accept or reject orders without penalty, connection and disconnection without restrictions, absence of exclusivity, power to subcontract or designate substitutes and inexistence of a rating or control system with effects on the allocation of services. None of those elements is a clause: they are operational features of the model. The lesson for any platform is that autonomy is proven with real functioning, not with the wording of the contract.
The symmetrical lesson for riders and their representatives is that the name of the model does not win the case. What matters is whether, in the documented practice, the algorithm rewards availability, penalizes rejections or concentrates allocation on the most obedient. When that happens, the presumption is reinforced; when the company proves the opposite, it can fall.
The contrast with 2020 is not a contradiction
The Chamber recalls the judgment of the Fourth Chamber of the Supreme Court of 25 September 2020, which declared the employment status of Glovo's riders, and places the dispute on the correct ground: determining whether the subsequent modifications of the model alter the elements that sustained that classification. It is the serious reading of the doctrine. The 2020 judgment analyzed the 2020 model; it did not canonize a company nor condemn a technology. Each file is decided on the model in force and the evidence taken in it.
This has an uncomfortable consequence for everyone: the classification of the same rider can change over time and territory, because the model changes. For firms advising platforms, the continuous documentation of autonomy features becomes evidentiary compliance. For those representing riders, investigating the real functioning of order allocation is the central piece of the claim.
The contentious channel is not the labor channel
The Extremadura case arrived through the contentious-administrative channel, against resolutions of the General Treasury of Social Security on registrations and deregistrations. It is the terrain of the Inspectorate and contributions, different from the rider's individual claim seeking recognition of their employment relationship before the labor jurisdiction. A ruling in one channel does not bind the other, and the two coexist with criteria that do not always coincide, as the case law of recent years has repeatedly shown. Whoever reads this judgment as a definitive closure of the debate, in either direction, is wrong: the resolution is neither final nor does it exhaust the procedural map.
The cassation appeal before the Third Chamber of the Supreme Court is open, and with it the possibility that the debate moves up to the national level. If cassation arrives and prospers, there will be contentious doctrine of national scope on the Flex model; if not, the ruling will remain one more territorial precedent in a diverse mosaic.
The underlying fact: the Inspectorate no longer wins by default
For years, Labor Inspectorate actions against delivery platforms moved on favorable terrain: the presumption reversed the burden and the company rarely managed to document autonomy with sufficient precision. The Extremadura ruling marks an evidentiary turning point, not an ideological one. When the company arrives at trial with a redesigned model and documentary evidence of each autonomy feature, the administrative resolution supported only by the presumption is left exposed. The Inspectorate will have to prove facts (real penalties, effective controls, de facto exclusivity) and not limit itself to invoking the rule.
It is a foreseeable evolution of any legal presumption: the first moment is shock for the presumed party, the second is adaptation. The platforms have adapted. The debate over whether that adaptation is substantive or cosmetic is, precisely, what the courts will keep resolving file by file.
What each side should note
The platform: the model is litigated with data, so the terms of use, allocation records and penalty policies must sustain the autonomy narrative that will later be defended before the Inspectorate and the judge. Changing the model without documenting the change is litigating blind.
The rider: the presumption plays in their favor, but the file is won by proving the facts that belie it in practice: sanctions for rejecting orders, connection incentives, degradation of allocation. The individual claim before the labor jurisdiction remains open regardless of what the contentious channel decides.
The administration: the Inspectorate acts on the model of a given moment; resolutions based on uncontrasted presumptions are vulnerable, as this ruling demonstrates.
What neither side should forget
The public conversation about riders usually stays at the label (employee or self-employed?), but the rider's life is decided in the details: who bears the risk of days without orders, who pays sick leave after an accident, who contributes and on what base. The legal classification is the instrument to answer those questions, not an end in itself. That is why the serious discussion is not only evidentiary: it is about what social protection corresponds to work organized by app, whatever its formal fit. That question belongs to the legislator and to collective bargaining; the courts only apply the existing categories to the model before them.
A judgment that invites prudence
The media temptation will be to present the ruling as the end of rider employment status or as an isolated error about to be corrected. Neither reading is serious. What there is is a non-final judgment, in a specific channel, on a specific model, which confirms the underlying rule: the employment presumption is strong, but it is a presumption, and presumptions live on evidence to the contrary. The debate continues where it always was: in how the allocation of each order really works. Everything else is noise around a question that only the evidence can settle.
FAQ
Does this ruling mean Glovo riders are no longer employees? Not in general. The judgment, not final, declares the inexistence of an employment relationship in the case tried under the Flex model proven in that file.
Is the employment presumption still in force? Yes. The judgment does not question the presumption, but considers it rebutted by the evidence taken in that case.
Can the Supreme Court review it? Yes, the resolution admits a cassation appeal before the Third Chamber.
Sources
Opinion piece anchored in the official judicial news of the Poder Judicial of 12 November 2026, verified on 14 November 2026. The note does not include the full text of the judgment; the analysis is limited to what was institutionally communicated.
Fuentes
- Source. Accessed 14 November 2026.
The specific data in this article come from the official sources cited.


