Artificial Intelligence’s integration into the office has sparked significant academic and public debate. Most discussions have focused AI’s impact on employment. Meanwhile, it’s consequences on workers’ psychology and behaviour has been largely ignored. A new study, titled “Deployment of algorithms in management tasks reduces prosocial motivation” by Armin Granulo, Sara Caprioli, Christoph Fuchs, and Stefano Puntoni, addresses this less-explored dimension. The research investigates how the use of algorithms in “management tasks” (e.g., evaluating workers’ performance) affects employees’ prosocial motivation (the desire to protect and promote the well-being of others), a key aspect of workplace productivity and social interactions.
The findings reveal a consistent pattern: employing algorithms for management tasks, as opposed to human managers, leads to a significant reduction in employees’ prosocial motivation. This effect is attributed to increased objectification of coworkers, where they are perceived more as tools rather than humans with emotions and individuality. The paper demonstrates that this negative effect occurs even when algorithms and human managers work together and varies depending on the type of management task performed by the algorithms. These are invaluable lessons as we increasingly turn to AI in the office.
Key takeaways from the study
The use of algorithms in management significantly diminishes employees’ willingness to engage in helpful and cooperative behaviors.
The deployment of algorithms leads to a perception of coworkers as tools or resources rather than as individuals with emotions and personal qualities, contributing to the decline in prosocial behavior.
Even collaborative management involving both algorithms and humans does not fully mitigate this effect.
The negative effects on prosocial motivation vary depending on the specific management tasks performed by the algorithms, indicating that not all algorithmic interventions are equally impactful.
What do we do with that information?
In order to learn from this type of research, companies can put in place palliative actions, whether or not they plan to integrate algorithms to their process immediately, or in the near future.
Promote human Interactions
Encouraging more direct human interaction and collaboration in the workplace can counterbalance the depersonalizing effects of algorithmic management. This involves fostering an environment where employees engage more with each other on a personal level (recommending more in-person discussions, for example). There are benefits to “return-to-office”, and we shouldn’t be blind to them.
Find the right balance
Integrating human insight with algorithmic efficiency, especially in areas requiring emotional intelligence and nuanced decision-making, can mitigate the negative impacts. As highlighted above, mixing algorithms with humans to make decisions changes little… but pro-social behavior may not be something to systematically strive for in a company setting. So we should ensure human managers always handle more sensitive or complex interpersonal tasks, while remaining open to a little algorithm support in specific places.
Develop ethical guidelines
Developing and enforcing clear regulations and ethical standards for AI use in management roles can help ensure these technologies are used responsibly and with consideration for employee well-being.
Meanwhile, promoting transparency in how AI tools are implemented and involving employees in the process can help in creating a more accepting and understanding environment. This includes educating employees about the purpose and function of AI tools in their work.
Too soon to draw conclusions
Though this is a fascinating look at one possible future, there’s plenty to be wary of in the study we are discussing.
The paper primarily uses correlational studies and controlled experiments, which may not fully capture the complexity of real-world organizational dynamics.
The generalizability of the findings could be limited due to the diversity of workplaces and cultural differences.
Further research is needed to explore long-term impacts and the interaction between algorithmic management and various organizational cultures.
Finally… we need to wonder what companies exit for. Of course, it’s better to have employees be helpful and cooperative. But is it always necessary?
As scientist often say… more research is needed.
While the adoption of AI in management shows efficiency gains, it raises concerns about the erosion of positive social behavior in the workplace. The findings suggest a need for a balanced approach to integrating AI into management practices, considering the psychological impact on employees.
There is hope that with thoughtful implementation and regulation, the benefits of AI can be harnessed while preserving the social fabric of the workplace.
Good luck out there.
