The same tension — trust in data, but suspicion of opacity — resurfaces at the boardroom table, even though the questions there concern strategy rather than working-time tracking. A survey conducted in spring 2026, answered by 323 leaders in expert and knowledge work, found that 60 per cent perceive algorithmic management to be present in their own workplace to at least some degree, and 91 per cent identify its presence in at least one named leadership task (Lahtinen et al., 2026).

International research points in the same direction: a recent review synthesising 172 peer-reviewed studies shows that algorithmic management is entering a new phase; the centre of gravity is shifting from platform-economy transport and gig work (Uber, Wolt, Amazon, for instance) towards expert organisations and their strategic decision-making (Lippert et al., 2026). Haaga-Helia’s research, then, isn’t trailing the international conversation — it lands squarely at the turning point the whole field has identified.

The figures climb once we move into strategy work. According to Haaga-Helia’s survey, roughly 30 per cent of leaders believe a machine could outperform humans at strategic planning and forward-looking thinking (Lahtinen et al., 2026). Anna Lahtinen, senior researcher and project manager of The Machine as Manager project, describes the finding as surprising. She expected the machine to do well on number-crunching tasks, but hadn’t anticipated strategic thinking emerging quite so strongly (ibid.).

Algorithmic management’s clearest selling point appears to be fairness. As many as 73 per cent of leaders see it as a positive that the machine bases its decisions on data rather than personal relationships. On the other hand, AI’s potential to boost engagement or workplace wellbeing was rated as weak (ibid.). The results also suggest that gender, age or sector doesn’t meaningfully relate to trust in AI. 

Frequency of use, however, does raise favourability towards it. The same pattern already showed up in the 2025 knowledge-worker survey (Asikainen & Lahtinen, 2025), and internationally, especially among those who have previously experienced human management as unfair (Moritz & Wehner, 2026).

This is worth treating with a healthy dose of caution, though. Rosenblat and Stark (2016) showed, in their study of Uber drivers a decade ago, that perceived fairness can be something of an illusion. An algorithm looks neutral because it doesn’t play favourites by name, but the power structure behind it — who sets the metrics, and on whose terms — still favours the system’s owner, not its user. There’s a second phenomenon at play here too. Taina Bucher (2018) calls this “pacifying the algorithm”: employees gradually learn to adjust their behaviour to what they believe the system is measuring, regardless of whether that reflects the actual quality of their work.

But more interesting than fairness is what leaders are not prepared to hand over to the machine. More than 80 per cent would rather work with an algorithm that’s 75 per cent accurate but comprehensible than one that’s 95 per cent accurate but a black box, a system whose decision-making process isn’t transparent to the user or the organisation (Lahtinen et al., 2026). 

The distinction matters, because as Ananny and Crawford (2018) point out, seeing an algorithm isn’t the same as understanding it, and simply having access to a system’s logs doesn’t in itself explain why it arrived at a particular recommendation or conclusion. Accuracy alone isn’t enough, then — leaders also want to understand what’s actually going on, even at the cost of a less accurate result.