Most organisations approaching AI adoption are answering the wrong question. They are asking, "How do we get our people fluent in these tools?" The harder, more consequential question is the one almost nobody is asking: "Who are we as a team now that AI is part of how we work?"
That second question is an identity question. And in the social identity tradition that runs from Tajfel and Turner through to the contemporary identity leadership work of Haslam, Reicher, Platow and Steffens, identity is not a soft variable. It is the strongest available predictor of trust, effort, voice, ethical conduct and willingness to follow.
The Microsoft 2026 Work Trend Index, with twenty thousand respondents across thirty one countries, makes the point in numbers. Across the twenty nine factors it measured, organisational variables (culture, manager support, talent practices) account for sixty seven per cent of the explained variance in AI impact. Individual mindset and behaviour account for thirty two per cent. Organisational AI culture is the single strongest predictor. The team's identity, its sense of who we are and what we are for, is doing more work than any individual's tool fluency.
What we mean by identity
Social identity is the part of a person's self concept that comes from membership of a group. When that membership is salient, people think, decide and act on behalf of the group, not just themselves. A team with a clear, shared identity makes faster decisions, shares information more freely, recovers from setbacks more quickly, and resists drift toward decisions that would embarrass the people they believe themselves to be.
A team with a fuzzy or contested identity does the opposite. It hesitates. It hedges. It defers to whoever has the loudest voice or the most plausible artefact. In an AI augmented context, that artefact is increasingly an output from a model.
Why AI changes the identity question
Three things happen when AI enters a team's workflow that change the identity calculus.
First, the team's information environment changes. Previously, information came from named human colleagues with track records the team understood. Now, a meaningful share of inputs come from systems whose reasoning the team cannot inspect. The team has to decide, often without discussion, how much weight to give those inputs.
Second, the team's division of labour changes. Tasks that used to define the work, the analysis, the first draft, the synthesis, are increasingly being done by AI. The team is left with a residual set of activities that may or may not match what the team thinks it is for.
Third, the team's social fabric changes. Yan and colleagues, in a 2025 study with nine hundred and five participants, showed that AI personas reshape team behaviour across analytical, creative and ethical tasks even when participants do not detect AI involvement. The team's identity is being formed by default, by the design choices of software vendors, in the absence of an explicit conversation.
Build this capability in your organisation
Growth Performance designs bespoke programmes that turn insight into measurable behaviour change.
The augmented us
The Augmented Us is a framework for doing that conversation deliberately. It is not a tool fluency programme. It is an identity leadership programme for teams operating in human AI hybrid configurations.
The work is organised around six stages. Readying wakes leaders up to the silent identity work AI is already doing. Reflecting maps the team's identity ecology, including the AI tools and agents in it. Representing articulates what the team exists to do, and what AI is for inside it. Realising pairs every priority with an explicit human AI division of labour. Reinforcing installs a monthly ritual that surfaces drift before it hardens. Recalibrating provides the quarterly reset as AI capability and team composition shift.
Four artefacts carry the practice. The augmented social identity map. The team charter for augmented work. The human AI division of labour tool. The monthly continuity ritual.
Why this is leadership work, not technology work
The temptation, when AI is the change driver, is to make the work technical. Procure better tools. Run more training. Stand up another centre of excellence. None of that addresses the identity question, and the identity question is what the evidence says is moving the dial.
The leader's job in an augmented team is not to be the most fluent operator of the tools. It is to hold the team's sense of who we are and what we are for steady enough that the team can experiment, disagree, recover and learn without losing itself in the process. That is the work that tool training cannot do, and it is the work that distinguishes the teams that are getting somewhere with AI from the teams that have spent two years in pilot.
Try this
Ask your team a single question this week, in a meeting, with no preparation. "If a new joiner asked you what AI is for in this team, what would you tell them?" Listen to the variance in the answers. The variance is the work.
Get the full framework. This article draws on The Augmented Us white paper, the evidence based primer with the full six stage practice, the four practitioner artefacts and the research base. Download the white paper or take the team diagnostic.