What lessons can we learn from the pioneers of AM?

Algorithm-based work: when digital management undermines workers

The rise of digital platforms has deeply transformed working conditions in many sectors, particularly taxis and delivery services. As soon as the first forms of algorithmic work appeared, researchers observed a shift in decision-making power: companies no longer relied on human supervisors but on automated systems that evaluated, directed, and sometimes even excluded workers. While this development has been presented as a technological advance bringing flexibility and autonomy, workers’ experiences reveal a very different reality.

Algorithmic management comes with a series of structural problems that profoundly affect well-being, safety, and dignity at work.

Initially, it appeared that algorithmic systems introduced a form of dehumanised management that did not consider the actual conditions of work. Platforms began collecting vast amounts of data on workers, continuously analysing their activity and imposing intensive workloads without offering any support in return. This constant surveillance created a sense of exhaustion and pressure, especially since workers have access to only a tiny fraction of the information collected about them. Thus, rather than giving workers more control, the “algorithmisation” of work has mainly increased their vulnerability to opaque rules and to automated management focused on economic efficiency rather than human health.

As these practices took root, another major problem became apparent: platforms built an incentive system that resembled a manipulative gamification mechanism.(1) By incorporating elements borrowed from video games, such as levels, variable rewards, badges, and weekly “challenges,” they encourage workers to work longer hours and accept low-paying jobs, at the risk of their own health. Drivers and delivery workers find themselves caught up in a dynamic where they must continually “play” in order to earn more, even though the rules of the game are constantly changing, and the best bonuses are often reserved for newcomers. Far from being entertainment, this gamification becomes a sophisticated control tool, capable of pushing workers to try to push their own limits while giving the illusion of voluntary choice.

At the same time, information asymmetry is one of the most critical aspects of algorithmic management. Platforms know everything there is to know about their workers: their break times, their movements, their acceptance rates, their hourly earnings, their speed, and even their preferred parking areas. In contrast, workers have only a fragmented view of the information they need to make informed decisions. They generally do not know the destination of a ride before accepting it, do not understand the criteria for assigning rides, cannot verify the accuracy of compensation calculations, and have no visibility into the mechanisms that influence their ratings. This asymmetry creates a deeply unbalanced power relationship, in which workers must navigate blindly despite the risks of economic loss or physical danger. The promise of a balanced relationship between “independent partners” seems all the more hollow when decision-making power remains concentrated in the hands of an opaque, inaccessible technical system.

 

—

_{[1] Gamification : integrating fun elements into work activities}

Finally, one of the most insidious effects of algorithmic work is the radical individualisation of labour, which leads to organised isolation. Unlike traditional workplaces, where employees can interact, help each other, or mobilise collectively, platform workers operate alone, in their vehicles or on the street, constantly on the move and without a shared space. Algorithms assign tasks individually, evaluate them individually, reward them individually, and punish them individually. This systematic fragmentation not only undermines workers’ ability to build solidarity but also prevents the emergence of collective strategies to address structural injustices. Isolation thus becomes an integral part of the platform economy model, limiting the possibility of protest and weakening workers’ collective voice.

In short, algorithmic management is fundamentally altering the nature of work, imposing intensive surveillance, incentive-driven manipulation, structural opacity, and social isolation. Despite platforms’ claims of innovation and flexibility, workers experience declining conditions and diminished agency. This contrast highlights an urgent need to scrutinise algorithmic management and prioritise workers’ rights, dignity, and health in public and union debates.

Gamification is coming to public services: opportunity or managerial excess?

Gamification is no longer just for private companies. It is now entering public administrations, where it helps to engage citizens and employees through reward, challenge, and competition mechanisms. This shift reflects a broader transformation in managerial practices as public services become digitised.

Initially, this approach appears to appeal to many institutions, particularly in professional training. In France, for example, organisations are increasingly using gamified platforms to improve employee motivation and learning attitude, even though the scientific evidence remains mixed in comparative studies(1). But gamification now extends beyond training. Welcome to managerial gamification! It is finding its way into recruitment, internal innovation tools, and even change management processes within government agencies. Specialised journals on public transformation document the creation of role-playing games for interviews, “serious games”(2) for training in service design, and even fun tools designed to speed up certain public decisions.

This development presents significant challenges for trade unions. Gamification, despite its playful appearance, is a deliberate tool for shaping workplace behaviour. It promotes individual performance and competition, directly clashing with the values of public service, such as cooperation, equal treatment, and community. Furthermore, it operates as a subtle means of managerial control, intensifying workloads and conditioning practices without genuine debate.

Gamification can assist learning and engagement, but should never replace genuine professional recognition or mask management practices that erode employee rights. As public services face transformation, safeguarding meaningful work and professional values beyond superficial digital incentives is crucial.

 

 

—

_{[1] Mourad Bofala. Contribuer à l’apprenance des collaborateurs : le rôle de la gamification. Gestion et management. Université Pascal Paoli, 2022. Français. ⟨NNT : 2022CORT0004⟩. ⟨tel-03895804⟩}

_{[2] Ndlr : « serious game » : A digital game designed primarily not for entertainment, but for learning, training, awareness-raising, or professional simulation. It uses game mechanics (rules, scenarios, feedback) to facilitate the acquisition of skills or knowledge in an educational, organisational, or managerial context.}

Putting the lessons of algorithmic management into practice

Drivers, delivery workers, and micro-task platform workers were the first to experience algorithmic management, and their experiences now serve as a valuable warning to the entire world of work. In a 2019 report on “Digitalisation and public services: a trade union perspective,” Public Services International (PSI), in collaboration with the Friedrich Ebert Foundation, also emphasises that the social consequences of technologies depend entirely on how they are implemented, regulated, and negotiated, as well as on the ability of trade unions to influence their governance. The issues identified in pioneering sectors are now well known. They can be classified into two main risk categories. Those that pose a risk to rights:

• opacity of algorithms, creating unilateral power for employers;
• disguised deregulation, where technology is used to circumvent collective protections.
• information imbalance, entirely to the advantage of the platform;

and those that pose a risk to health and/or working conditions:

• lack of control over data, used for monitoring, evaluation, or punishment;
• isolated and individualised work, which weakens collective organisation;
• manipulative gamification, transforming incentives into tools for intensifying work;

In 2023, PSI launched a unique tool available in French, English, and Spanish: the Digitalisation Negotiation Portal, the first global database of actual negotiation clauses, framework agreements, union advice, and international resources to help unions negotiate all issues related to the digital transformation of work. Designed as a strategic resource centre, this portal gives unions access to content organised by theme—data rights, digital tool governance, teleworking conditions, AI and algorithms, health, safety, inclusion, worker consultation—and allows them to draw on clauses already negotiated in other countries or sectors. It is therefore an essential tool for regaining control in the face of technological developments that are profoundly transforming the organisation of work.

The Portal offers clauses regulating algorithmic transparency, allowing unions to demand information on automated decisions, similar to the struggles waged by ride-hailing drivers to understand the mechanisms of ride distribution. Pioneers have shown that technological opacity creates arbitrary power; the ISP responds by citing examples of clauses requiring prior consultation with agents or the explainability of AI systems. But it also makes recommendations to unions, such as avoiding restrictive language that imposes limits on tools and technologies, as well as on how they are used at work. “It is important to be explicit about how technologies and tools will be used, but it is equally important to specify how they will not be used. Many collective bargaining agreements include language on acceptable uses of technologies and impose limits on their use. With regard to restrictions, unions will ensure that technologies are not used for surveillance or punishment.”

Unlimited digitisation, cascading psychosocial risks

With regard to health and safety issues, Le Portail recommends a comprehensive approach to psychosocial risks (PSRs) that arise directly from the introduction of digital tools, AI, and algorithms: increased pressure linked to monitoring tools, risks induced by new work organizations (including teleworking), mental overload linked to information management (digital infobesity), and the effects of digital devices on stress and well-being.

The PSI emphasises the prevention of risks related to information overload, increased surveillance, the pace imposed by digital tools, and new work environments, particularly teleworking. These risks are addressed in Section 7 of the Portal, which provides concrete references for requiring impact assessments, protective measures, appropriate training, and measures to reduce stress or psychosocial risks. Health and safety must remain central obligations in digital transformation processes. Collective bargaining remains essential to regulate these technologies and provide lasting protection for public service workers.

In public services, as elsewhere, certain digital tools exacerbate psychosocial risks, particularly those that require constant hyperconnectivity, such as instant messaging (WhatsApp groups, Signal, Teams, Yammer, etc.), notification systems, and collaborative platforms (such as SharePoint), which foster a culture of urgency and constant availability, leading to mental overload and chronic stress. Added to this are management tools that dictate the pace and organisation of work, reducing employee autonomy and increasing the pressure associated with automated performance, a phenomenon widely observed in highly digitised environments. Finally, digital surveillance devices (cameras, keyboards, presence or motion sensors), geolocation devices (GPS trackers, geolocation applications, AirTags, etc.), and activity tracking amplify the feeling of constant control and deteriorate the psychological climate at work, while intensive teleworking, structured by these same tools, increases isolation and blurs the boundaries between private and professional life.

Conclusion – Digitalisation is not inevitable

Digitalisation is not inevitable, nor is it an unavoidable technological destiny to which workers must unconditionally adapt. It is a collective choice, a social construct that can be discussed, negotiated, and regulated. The experiences of platforms clearly show that unregulated deployment leads to surveillance, isolation, stress, and loss of control. But they also show the opposite: when workers take back control, when unions impose transparency, human oversight, data safeguards, and limits on the use of AI, the algorithm ceases to be a threat and becomes a tool again. The Bargaining Portal developed by PSI provides the means to transform these lessons into concrete clauses, enforceable rights, and effective protections. At a time when public services are investing heavily in digital technologies, the question is not whether to accept or reject these tools, but how they will be used, in whose service, and within what limits. Taking back control of algorithmic management means remembering that democracy at work does not disappear when code replaces a manager: it simply has to reinvent itself. In the public service, technology is a tool. Work, on the other hand, remains a human endeavour.