At the 2026 CESI Summer Days conference in Brussels, Samuel Goodger, Policy Analyst at the European Policy Centre, stepped in to deliver a workshop on Algorithmic Management and AI at Work, on behalf of Prof. Carmelo Cennamo, Program Director at Reshaping Work. This blog post explores key session themes, contextualising them within discussions on the EU’s upcoming Quality Jobs Act.
By Samuel Goodger and Carmelo Cennamo
Amid the rise of generative AI, algorithmic management – the use of digital systems to monitor and allocate tasks, among other functions – is increasingly diffusing across the workplace, far beyond the gig economy. Bringing with it a new form of workplace governance, decisions once made by managers are increasingly obscure, and harder than ever to challenge. This begets several questions. In an AI-powered economy:
- How must workers and managers navigate sophisticated algorithms?
- How should workflow be reconfigured and allocated between AI algorithms and human to preserve and augment labor skills and competitiveness, instead of substituting them?
- What consequences will this have for the quality of jobs, how we value labour, overall (firm and labor force) competitiveness, and democracy at work?
To set the scene, the workshop examined algorithmic management under four dimensions – context, agency, interaction and outcomes – before exploring use cases and discussing how algorithmic management changes managerial functions.
We applied then a critical lens to the positive and negative possibilities enabled by this transformation. For example, AI can increase satisfaction by providing personalised attention, and real-time feedback; on the other hand, opaque algorithmic systems can increase power asymmetry, by increasing uncertainty.
One particularly interesting area is on skills: algorithmic management and AI can both lead to stagnating capabilities due to de-skilling, but it can also boost learning and skills development. Which direction change manifests is partly a question of complementarity – whether AI augments a worker’s tasks or absorbs them. Unsurprisingly, this line falls unevenly across occupations, sectors and demographics.
What does this mean for the future of work, especially in the context of the new regulatory framework that will be introduced through the Quality Jobs Act? The upcoming legislation aims to ensure protection against data exploitation, boost fairness, strengthen social dialogue and improve labour rights in the EU labour market, against an incessant tide of crisis and change on multiple fronts. With AI increasingly widespread, 84% of Europeans believe its adoption at work must be carefully managed – and soon. As such, we propose a five-axis framework to properly frame the discussions above and ensure algorithms and AI lead to progress, rather than backsliding.
1. Autonomy
- Can workers influence how work is allocated, paced, measured, and evaluated?
2. Fairness
- Are decisions explainable, non-discriminatory, and context-sensitive?
3. Accountability
- Who is responsible when an automated decision harms a worker?
4. Voice
- Are workers and trade unions informed, consulted, and able to negotiate rules before deployment?
5. Personal development
- Who is responsible for workers’ career progression, ways to enable, and track it?
Beyond extending the Platform Work Directive’s worker protections to the wider economy, the Act should ensure several conditions:
- the right to know,
- to human review and contestation,
- to collective involvement, as well as
- algorithmic risk assessments,
- and clarify prohibited or high-risk practices (in line with the AI Act).
Without such provisions, the Act risks ignoring persistent – and widening – gaps around accountability, democracy and fairness.
These discussions resonated strongly with the audience of trade unionists, researchers and policymakers, who brought examples of conveniently unclear liability around AI decisions, increased AI-driven stress, or described the situation as a worrisome ‘collapse of nuance’ in the workplace.
Without careful, participatory governance at all levels, the tidal wave of increasingly capable AI may bring massive changes to our agency at work, the skills we use, and our goals, which risk AI-linked devaluation of work, or mass unemployment, rather than enhanced human capabilities and firm and workforce competitiveness.
Increasingly, such pressures fall hardest on middle-income earners, who have historically financed our welfare states. This makes job quality a question of labour-market resilience and competitiveness, as much as of workplace rights.
Importantly, this discussion should not lead us to reject change, but to imagine how to change it for the better – using AI to improve our health, reduce workload, or move on to more meaningful work. Beyond ensuring adequate social and labour protections remain in place, we must fight for guardrails and accepted uses for AI, enforce rules against tech oligopolies and, ultimately, ensure all democratic voices are heard in this transition. As the workshop put it: the Quality Jobs Act should not block innovation, but define the conditions under which AI at work supports productivity, without undermining human autonomy, dignity, and fairness.