We present the position of the Zentrale group regarding the planned transposition of the Platform Work Directive into the Polish legal system. The Ministry of Family, Labour and Social Policy has published a draft “Act on Performing Work via Digital Labour Platforms,” which is intended to be the cornerstone of this implementation. We are not labour law experts, so our position may have shortcomings. We also acknowledge that this is merely a draft, which could change completely during parliamentary debate. Ultimately, it could even be rejected by the President.
We believe that the scale of exploitation on these platforms is immense and widespread. It is high time for the authorities to finally take action.
#zentrale_dyrektywa
Platform Work Directive (PWD).
GENERAL INFORMATION ABOUT THE DIRECTIVE:
The legal act discussed here is known as the Platform Work Directive, though its official title is Directive (EU) 2024/2831 of the European Parliament and of the Council of 23 October 2024 on improving working conditions in platform work. As the full text was published on 1 December 2024 and each EU country has exactly two years to implement it, it is set to become part of the Polish legal system by 2 December 2026.
Work on the Directive began as early as 2021; at a very early stage, the Zentrale group was invited to participate in open consultations regarding its content.
We recognize that the adopted text of the PWD represents a compromise between the competing interests of various political and capital groups from different countries. It is also universal in scope, applying to everyone working for platforms—not just those involved in delivery or passenger transport services. However, given the nature of our own activities, our critique is framed from the perspective of individuals working as delivery personnel for platforms.
PLATFORM WORK AS PART OF SOCIAL LIFE.
Currently, so-called “platform work” consists of very simple tasks that do not require specialized education; precisely for this reason, platforms easily capture this segment of the labor market. The tasks are so simple that, in principle, neither formal education nor training is required. Consequently, such work is usually very low-paid, regardless of whether it is performed through a platform or independently. However, it is precisely these platforms that—by testing the limits of the labor market’s endurance—drive wage levels well below the minimum wage. Consequently, the people who opt to participate in this modern form of exploitation (worldwide) are generally the most desperate social groups or individuals who view the work as merely a stopgap measure from the start. Yet, this is changing. Having saturated the sectors involving the simplest tasks, platforms are now entering new, more complex sectors—though still only the lowest-paid ones (e.g., hospital nursing or cleaning services). High levels of exploitation are evidenced by constant worker turnover. However, because this turnover persists, platforms essentially enjoy a steady supply of new workers, allowing their operations to continue without interruption.
ECONOMIC SCALE OF THE PLATFORM MARKET
(BASED ON DELIVERY PLATFORMS ONLY)
UBER – 2025 revenue of approx. PLN 205 billion – equivalent to Poland’s entire annual military budget (PLN 200 billion).
Delivery Hero (Glovo) – PLN 63 billion – equivalent to annual spending on the “800+” social benefit program (PLN 62 billion).
Just Eat Takeaway (Pyszne.pl) – PLN 21.9 billion – equivalent to the annual cost of indexing all retirement and disability pensions (PLN 22 billion).
A common feature shared by all the aforementioned platforms is their failure to report a profit. Despite Uber having operated for over 15 years and maintaining a massive scale of operations, it still does not show a profit (and therefore pays no taxes). Thus, their operations are not only detrimental to the labor market but also harmful to the economies and national budgets of the countries where they operate (we discuss this further in our other publications).
POSSIBLE SCENARIOS FOR IMPLEMENTING THE DPP
The content of the DPP is so general that it allows for considerable latitude in interpretation during transposition into national legal systems. In our assessment, the result will be that implementation could be highly pro-worker in some countries while maintaining the status quo in others—all within the same European Union. How it is implemented in Poland is entirely a political decision made by those in power.
KEY ISSUES ADDRESSED IN THE DPP (analysis of its main premises)
1. Employment status: The Directive introduces a presumption of an employment relationship. If the relationship between the platform and the worker exhibits elements of direction and control (e.g., the platform determines remuneration, supervises work, or influences task organization), the agreement between them is deemed to constitute an employment relationship. The DPP reverses the burden of proof—it is the platform, not the worker, that must prove it is not an employer.
2. Algorithmic Regulation: A defining feature of platform work is the use of an algorithm (a computer program) to manage work. Algorithms play a dominant role in platform operations; in some cases, the algorithm makes over 90% of decisions entirely on its own. The proposed Directive (DPP) requires platforms to disclose the content of their algorithms. Workers must be informed about how automated monitoring and decision-making systems operate (e.g., regarding task allocation, pay determination, or account suspension). Additionally, the DPP stipulates that key decisions must be made by humans.
3. Personal Data Protection: Under the DPP, it will be more difficult for platforms to collect worker data due to the economic dependency involved. It prohibits the processing of certain particularly sensitive data (e.g., emotional state, private conversations, or worker location data outside of working hours).
4. Strengthening the role of workers and public authorities in relations with platforms. The DPP allows for the imposition of reporting obligations on platforms regarding their activities to the relevant supervisory bodies. It also mandates that platforms facilitate free communication among the individuals working for them.
GENERAL CRITICISM OF THE DIRECTIVE CONCEPT ITSELF
regarding the points mentioned above.
re 1.
The DPP grants governments significant leeway in interpreting its provisions. Each government will make a political decision regarding the definition of “direction and control” over the work and the worker. Some countries list specific criteria that, if met, classify an individual as an employee (e.g., setting working hours, performance evaluation, determining remuneration). Others will establish a different set of criteria. It is their political decision that will determine whether the criteria are easy or very difficult to meet. Our observations indicate that some countries leave even these issues for the courts to resolve. Others employ a mixed system: they set certain criteria, but leave detailed decisions to the courts. Consequently, such implementations will change nothing. A lawsuit filed against a platform by a worker who intended from the start to work for only a month or two will never happen. Furthermore, we believe that no worker earning less than the minimum wage would choose to launch a legal battle against a platform with billions in revenue. Obtaining support from trade unions is also unlikely. Currently, no trade union confederation has a section specializing in platform work, and the only group active in the Polish delivery sector—the OPZZ Konfederacja Pracy branch at Pyszne.pl—has effectively disintegrated.
re 2.
Proponents of “algorithmic transparency” overestimate the importance of this aspect. It is widely believed that understanding the algorithm’s operating mechanism allows one to avoid exploitation. In our view, this is a dangerous myth that distracts from the true essence of work: earnings. Here are a few of our arguments:
– An algorithm is a massive, constantly evolving computer program. The DPP does not specify how or how often the algorithm’s code should be disclosed. Should it be a printout? A spreadsheet? Raw code sent via email? These questions may seem trivial, but they are absolutely crucial, as the form of disclosure determines whether a reliable analysis is even possible. Merely organizing the provided data can consume vast resources.
– As mentioned above, the algorithm is constantly being refined and altered—changing from day to day, or even hour to hour. What is the point of knowing yesterday’s code when today’s is different?
– Who is actually going to analyze the code? It requires a team of IT specialists. Are trade unions supposed to foot the bill? Especially unions that do not even exist in this sector?
– And perhaps most importantly: if we do analyze the algorithm, what does the worker gain from it? For instance, suppose we reveal that the algorithm adds an extra 12% to a novice courier’s pay to incentivize them. How does this knowledge change things for the workers? What are they supposed to do with it?
The DPP stipulates that “key decisions must be made by a human.” But which decisions should be considered key to daily work? This is crucial, as it is precisely what the platforms will undoubtedly use in the event of legal disputes. Another point: what does “human intervention” actually mean in practice? Is it enough for an algorithm to select a worker for dismissal and for a human simply to click “approve”? Or is the human required to thoroughly review the justification and hear out the arguments of the worker facing dismissal? Let us remember: the so-called “support” staff for delivery platforms are simply call-center workers subject to the same algorithmic pressure. They are never specialized personnel. So, will exploited platform workers be subject to review by stressed, overworked call-center staff?
re 3.
The general concept of privacy protection is very sound. Platforms genuinely violate workers’ privacy on a massive and continuous scale. These are not isolated incidents, as platform representatives claim; numerous analyses prove this. The apps used by platforms can even record workers’ phone calls outside of working hours! There is also evidence that workers’ movements are tracked even on days when they are not working. All this information can be used to further algorithmize earnings. However, as with the previous point: what is the use of enhanced personal data protection if there is no standard for how the collected data is shared? Some platforms provide it in the form of electronic spreadsheets, while others provide multi-page PDF scans. Without easy, standardized access to this data, we cannot even dream of critically analyzing it. Furthermore, one never knows… …whether the material they are handing over is truly 100% of what they have collected.
re 4.
This point seems like a very sound move. However, it won’t change much. Employees are already creating communication channels of their own that employers know nothing about. Of course, the platforms themselves also create such channels (hand in hand with fleet partners), giving the impression that they facilitate open discussion. And the DPP will change that. But will it change workers’ earnings?
CRITIQUE OF THE POLISH IMPLEMENTATION OF THE DPP
According to the draft bill presented by the government to introduce the directive into the Polish legal system, a so-called presumption of an employment relationship is to apply in situations where management—or “worker control”—is present. Unfortunately, however, there is no definition of what this actually means; the act does not specify it. Consequently, the courts will have to define it. And as we noted earlier, no delivery worker will realistically choose to initiate a lawsuit. In the case of migrants working for platforms, the prospect is entirely illusory; no one is going to wait years for a verdict over a claim involving 300 PLN plus social security contributions. Such a complex case would be deliberately dragged out indefinitely by the platforms. Furthermore, courts unfamiliar with the specific nature of platform work might rely on definitions of concepts that do not apply to this type of work, thereby ruling in favor of the platforms.
Under the draft bill, a decision by the National Labour Inspectorate (PIP) confirming the existence of an employment relationship takes effect only from the date the decision is issued. While the PIP is required to file a lawsuit to establish the employment relationship for the preceding period, this constitutes a separate legal proceeding—meaning there are two proceedings in every instance. Additionally, this provision implies that the worker receives legal protection only from the moment the PIP decision is issued; without that decision, there is still no protection whatsoever. What happens if a worker requests a PIP decision and an inspection begins, but the worker is dismissed in the meantime? The bill stipulates that if a platform terminates the contract during the inspection, the date the inspection began is deemed the start date of the employment contract. However, practical experience shows that platform workers lead such unstable lives that they cannot afford to wait—for instance—three months for the National Labour Inspectorate (PIP) to issue a decision, followed by another three years for a court to uphold it. This is a complete legal fiction. We are convinced that this will become a dead letter—a law that no one ever actually uses.
The bill still fails to address the issue of fleet partners (i.e., intermediaries). While the draft introduces joint and several liability for the platform and the partner, this applies only to the scope of the contract between them. This liability is waived if the fleet partner fails to notify the platform about the worker. Consequently, if the partner is a shell company (with no assets), the worker will essentially still have no way to enforce a claim, even after years of holding a court judgment in their favor. A separate and truly outrageous issue concerns the sanctions for platforms outlined in the bill. The maximum fine for a platform is PLN 60,000—a truly symbolic amount for global corporations. Earlier in this analysis, we cited data on platform revenues; for Uber, a PLN 60,000 fine is equivalent to a one-grosz ticket for someone earning PLN 5,000 a month. Platforms will simply treat such low fines as a cost of doing business—one that is easier to bear than changing their business model or hiring employees on standard employment contracts. To be effective, sanctions should be tied to a platform’s revenue or scale of operations.
A completely separate issue concerns workers performing tasks as sole traders. The draft bill excludes these individuals from the scope of certain provisions; for example, the right to an explanation of an algorithmic decision does not apply to those classified as “business users.” This means that workers operating as sole traders (JDG) will not gain the rights resulting from the Directive’s implementation!
THE FUTURE POST-IMPLEMENTATION:
In 2026, the National Labour Inspectorate (PIP) is set to receive significant new powers. Consequently, PIP branches will likely be flooded with new cases. Furthermore, their mandate will expand further upon the Directive’s implementation. We are likely facing an overload and potential collapse of the PIP. Should this occur, the state administration would once again be discredited. Additionally, as we noted in our report a year ago, the PIP lacks the substantive expertise required for the specific nature of the platform labor market. Perhaps the situation has changed since then; we certainly hope so.
SUMMARY
The Directive and its implementation initially appeared to be a breakthrough in the fight against exploitation in this labor market sector. However, in its proposed form—at both EU and national levels—it risks becoming a dead letter. The fact that overburdened courts will decide workers’ fates, that platforms will face only token penalties, and that the issue of “fleet partners” is completely ignored means the burden of fighting for absolutely basic labor rights continues to rest on the workers themselves.
To ensure the Directive’s implementation in Poland is not merely a sham, truly bold changes are required: equipping the National Labour Inspectorate with real tools and immediately enforceable powers to issue decisions with retroactive effect, and linking penalty amounts to platform revenues. And, of course, extending protection to everyone who actually performs work—including those in forced or voluntary self-employment.
WHAT’S NEXT?
Experience teaches us that no provision in a directive or statute will work on its own unless there is real power backing the workers. The Platform Work Directive provides certain legal tools, but whether they are actually used depends on our stance. Neither platforms nor algorithms will voluntarily give up their profits. The coming months of implementing these regulations in Poland must be a time for building solidarity among delivery workers and ramping up pressure on those in power. After all, the government is once again simply leaving us to face labor exploitation on our own, pointing to court proceedings as the only solution.
And furthermore:
THERE ARE 2 MONTHS AND 18 DAYS LEFT TO IMPLEMENT THE PLATFORM WORK DIRECTIVE.