Speech

GenAI in teams: a new organisational dynamic is needed

The introduction of GenAI into strategic working teams profoundly changes the dynamics of interaction and accountability

by Tommaso Buganza and Daniel Trabucchi*

23 September 2026

Aggiungi Il Sole 24 Ore
ai preferiti su Google

Manager nell’era dell’AI: le competenze che servono

4' min read

Translated by AI

Versione italiana

4' min read

Translated by AI

Versione italiana

generative artificial intelligence, now widely used by businesses, is also making its way into group processes: meetings, project teams, and situations where people need to discuss an issue or make decisions. But what happens when GenAI becomes part of the group?

This is a pertinent question, because managerial work is often a collaborative endeavour. Strategic decisions, innovation processes, and the design of new services and business models arise from the pooling of expertise. Introducing GenAI into these contexts means changing the way the team thinks, coordinates its work and takes responsibility for its own decisions.

Team dynamics

We studied this phenomenon at the Platform Thinking HUB Observatory at the Politecnico di Milano, as part of a research project carried out with Elisa Farri and Gabriele Rosani, authors and international experts in AI and organisational transformation at the Capgemini Invent Management Lab. For five months, we followed 60 managers from 12 companies across different sectors. The teams, comprising three or four people, worked in person on innovation projects, and the research examined the interaction between team members and GenAI, observing the dynamics during workshops and analysing all conversations.

When we add a new member to a team working on a strategic project, we expect the various members to interact with them. Instead, we observed something different: as a rule, when a GenAI was introduced, one person would take control of the keyboard, formulate the prompt and read out the response, whilst the others looked on. The group appeared to be using the GenAI, but in reality the interaction took place mainly between one individual and the machine. Silence, acquiescence, loss of control over the project.

The aim of adding GenAI to the team was to improve the quality of discussion, the evaluation of alternatives and interaction, but the initial effect was the opposite: teams tended to accept a direction proposed by the machine too quickly, thereby reducing discussion and the quality of the solution. All the managers involved were frequent and advanced users of that type of technology, so the problem lay in a lack of organisational expertise.

Team interaction with Gen AI

So, how does a team interact with GenAI? What are the rules and best practices to follow? The answer emerged from an analysis of conversations: we need the ability to integrate GenAI as a discussion partner for the whole group, capable of enriching the discussion and supporting human judgement. We’ve called it ‘Human-AI Team Chemistry’: a new dynamic between teams and artificial intelligence, built through a few simple yet deliberate practices.

The team must first and foremost make itself ‘visible’ to the GenAI: it must clarify who is present, what their areas of expertise are, and what problem is being tackled. In a meeting with an external consultant, this step would come naturally, but with GenAI it is often skipped, with the result that the machine believes it is interacting with a single person.

GenAI should also be assigned different roles: a client, a competitor, a facilitator, a prototyper, a storyteller. It can help to develop a proposal, but also to challenge it, by interacting with the various team members in specific ways.

More depth, not just more speed

The trickiest part, however, is guiding the reasoning. GenAI quickly generates options, summaries and action plans: this speed is a double-edged sword, as there is a risk that the machine will end up managing the innovation process, rather than the team. Before sending a prompt, the group must collectively agree on what it wants to achieve and then take the time to evaluate the response, refine it, and decide whether or not to use it.

This is where you can see the difference between uncritically delegating the project and thinking things through more carefully, whilst retaining control. In the teams that achieved the most promising results, the value of GenAI became apparent precisely in its ability to broaden the range of options, to explore matters in greater depth – perhaps even taking more time: the group did not bring the discussion to a close too soon and retained the final say.

A responsibility for leaders

Human-AI team chemistry does not arise simply from the availability of the tool; it requires new collaborative routines. Leaders play a central role in this process, as they can decide when to bring GenAI into a meeting, in what capacity, and under what rules of interaction.

Distributing licences and encouraging individual experimentation is therefore only the first step. The next step is to plan collective and collaborative work with GenAI: to arrange specific opportunities for interaction, introduce pauses for reflection, review transcripts, understand how the team reasoned, and improve the quality of subsequent conversations.

Value does not arise from collaboration being replaced, but from its evolution. GenAI can join the discussion, but the quality of the outcome depends on humans and on the rapport the team is able to build with this new presence.

*Lecturers in Leadership and Innovation and Platform Thinking at the POLIMI School of Management

Brand connect

I prossimi eventi

Tutti gli eventi

Notizie e approfondimenti sugli avvenimenti politici, economici e finanziari.