The Identity Renegotiation

What AI Asks People to Give Up Before It Gives Anything Back

Issue 276, August 6, 2026

Regular readers of Ideas and Innovations may have noticed that this newsletter and The Human Factor Podcast have been circling around something for months. In The Algorithmic Mirror, and its sister podcast episode, we examined what AI reveals about how we actually think and decide and found that the discomfort was never really about the technology. It was about the reflection. In The Readiness Illusion, we tested the AI agent era’s claims against the evidence and found a large gap between what organizations communicate and what their people are prepared to absorb. In The Empathy Outsource, we asked what happens when the care in a message was real, but the author wasn’t, and discovered that the question destabilized something deeper than the practice of effective communication. In The Confidence Transfer, we followed organizational expertise as it migrated from people to systems and watched what happened to the people when their judgment wasn’t the first thing consulted. And in Transformation Theater, we named the elaborate rituals of change that produce everything except change, and identified professional identity as the terrain where the performance most reliably collapses into failure.

Each of those issues examined one side of the same object. Every one of those patterns- the mirror we avoid, the readiness we overstate, the empathy we delegate, the confidence we transfer, the theater we perform- converges on a single unacknowledged transaction that sits underneath AI adoption in every organization I work with or talk to. I am going to call it the Identity Renegotiation: the process of renegotiating what a person’s expertise, judgment, and standing mean when intelligent systems begin doing work that used to define them.

Every AI initiative conducts this negotiation whether its leaders and those across the workforce know it or not. The only real choice leadership has is whether to conduct it openly, with both parties at the table, or silently, with one party dictating terms and the other party expressing its position through what we too casually label resistance.

This issue takes the thread further than any of the previous pieces could individually, because the argument only becomes visible when the pieces are assembled. So let’s assemble them.

The Transaction Nobody Names

We talk about AI adoption as if it were a skills transaction. The organization provides tools and training, the employee provides effort and attention, and competence is the exchange. The framing is clean, procedural, and measurable, which is precisely why adoption programs look the way they do: license counts, training completion rates, usage dashboards, prompt libraries. Readers of Transformation Theater will recognize these instantly as activity metrics. They measure the performance of adoption rather than the fact of it, and they allow everyone involved to report progress while the actual transaction goes unexamined.

The actual transaction most are unaware of is an identity transaction, and its terms are considerably less comfortable. We are asking people to surrender three things they spent their careers acquiring.

The first is mastery. The expertise that took a decade or two to build, the pattern recognition, the synthesis, the ability to look at a situation and know, is precisely what current AI systems replicate most visibly. This means the people with the most invested face the largest write-down, an inversion of every assumption our organizations make about seniority, expertise, and security. We have spent generations telling people that deep expertise is the safest asset they can hold. The systems now showing up in their workflows revalue that asset in an afternoon.

The second is the currency of judgment. In most professional settings, compensation operates on two ledgers. The first is the salary. The second is the standing that comes from being the person whose judgment gets consulted, and it is the ledger people actually organize their working lives around: evidence that they contribute, that they are productive, that their work has meaning. When an algorithm’s recommendation arrives before anyone asks the veteran what she thinks, that currency devalues quietly and without announcement. The Confidence Transfer traced the organizational half of this problem, the atrophy of institutional expertise when the machine is trusted first. The personal half is sharper. The person still holds the judgment. The organization has simply stopped asking for it, and everyone in the room can see that it has.

The third is standing itself. Expertise confers a place in the room, a reason to be included, a basis for belonging. If the expertise is now ambient and instant, available to anyone with a login and the right prompt, the place in the room becomes negotiable. This is the existential question sitting underneath every adoption plan: if the system can do what took a decade to master, what is that person’s value now? No amount of training answers that question, because it is not a question about capability. It is a question about worth.

Those in attendance at my Bridge 2026 session last week expressed this new, multifaceted and deepening uncertainty in their responses to the questions I asked them to discuss with their neighbors and their need to find answers continued individually with me once the session was over. Middle managers struggling to find the new path, marketers struggling with the change in role and scope, others contending with organizational downsizing and having to take on more responsibilities as overseers of AI outputs. And all of them unsure and unsettled about where this is all heading.

Here is the challenge that makes this new negotiation so fraught. The surrender is demanded up front, and it is concrete. The compensation, the promised elevation to higher-order work, the partnership with the machine, the redeployment of judgment to more consequential terrain, is delivered later, if at all, and it is and will remain far too abstract for the foreseeable future. AI adoption isn’t a technology rollout with a training and acceptance problem attached. It is a negotiation in which one party is asked to pay first, to trust, to accept ambiguity, and to bring a significant amount of patience. That patience can really be described as surrendering to the unknown and unknowable.

The Write-Down of Mastery

Consider what the surrender actually feels like from inside, because today’s adoption plans are written without any consideration for what individuals are going through.

Moving from expert to novice in public is one of the most psychologically expensive transitions a professional can make. Adults will actively avoid situations that expose incompetence, even when they intellectually accept that the new approach is better. This is why your most experienced people are so often your most resistant, a pattern I have now watched across dozens of client organizations and which I described through the three senior pharmaceutical scientists in Season 2 of the Human Factor Podcast. Their resistance to an AI-powered clinical trials initiative had nothing to do with doubting the models. The models were good, and they knew it. Their resistance was the entirely rational response of people who understood, faster than their leadership did, that the initiative was conducting a negotiation over who they were, and that nobody had invited them to it.

The write-down operates on three dimensions, and each compounds the others. Professional identity fusion means that years of expertise have become inseparable from self-concept. I am the person who knows how to do this. When a system makes the knowing obsolete, it doesn’t merely change a process. It invalidates accumulated mastery, and it says so in front of colleagues. Competence threat means the daily experience of being visibly worse at the new way than one was at the old way, with no guarantee the gap will close. And status and belonging risk means that the social architecture of the team, who defers to whom, who is consulted, who teaches, is suddenly under revision, with no stated rules for the revision. We are in a period of atrophy. As people adopt AI, their critical thinking, their expertise, and their other skills, even the interpersonal, are atrophying and will continue to atrophy.

Gallup finds that employees experiencing disruptive organizational change report frequent burnout at nearly twice the rate of everyone else, 42 percent against 24 percent, and that 64 percent of them are watching for or actively seeking another job. We read those figures as change fatigue, and some of it is. But a meaningful share is something more specific: the exhaustion of conducting an identity renegotiation alone, in private, against a leadership that won’t admit the negotiation actually exists. Grief without a venue is expensive, and the expense shows up on exactly the lines our dashboards track last, if at all.

Why Training Cannot Fix This

The organizational psychology literature has been circling this problem for decades, well before AI gave it urgency and a real-time case study. The research explains why the standard adoption playbook in use keeps failing and will continue to fail.

Jennifer Petriglieri, writing in the Academy of Management Review in 2011, described what people do when an identity they value comes under threat. They don’t simply comply or refuse. They engage in identity work: protecting the threatened identity, restructuring it, or in some cases shedding it. The critical insight for leaders is that this work happens whether or not the organization creates space for it. Deny it a legitimate channel and it goes underground, surfacing as workarounds, quiet noncompliance, selective enthusiasm in meetings, and hallway skepticism. All feeding into an organization’s culture that is being changed in very unintended ways. Leadership often misdiagnoses these symptoms as a change management problem and responds with more communication and more training, which addresses none of it, because none of it was ever about information.

Herminia Ibarra’s research on professional transitions, developed in Working Identity, supplies the second piece. Identity does not change by decision or in response to an announcement. People act their way into new professional identities through experiments, provisional selves, and new relationships, and the process is iterative, social, and slow. The change of identity doesn’t respect quarterly adoption targets, because it can’t. There is no communications plan that installs a new sense of self.

William Bridges made the companion point a generation earlier, and readers of this newsletter will remember the frame from our discussions of the neutral zone. Every transition begins with an ending. The psychological work of the ending, the letting go of who we were, is the phase organizations most reliably skip, because endings are uncomfortable to sponsor and impossible to celebrate at a kickoff. We schedule the launch. We don’t schedule the loss. And then we are surprised when people seem stuck, when the actual situation is that we asked them to begin something new without ever permitting them to finish something old.

Assemble those three bodies of work and the conclusion is difficult to avoid. Training addresses capability. The renegotiation is about identity. Applying training to an identity problem is a category error, a technical solution applied to an adaptive challenge, and it produces the same result: the appearance of adoption while monitored, and reversion the moment attention moves elsewhere. The spreadsheets come back. The old workflow returns. Not because people are stubborn, but because the negotiation was never concluded, and an unconcluded negotiation defaults to the status quo.

The Contract Beneath the Contract

Every organization holds two contracts with its people: the written one covering compensation and role, and the unwritten psychological contract covering expectations, norms, and implied promises. Denise Rousseau’s research established that the unwritten contract is the more powerful predictor of behavior, and that its violation is experienced not as change but as breach. Breach is the right word, because the felt experience is betrayal.

The Identity Renegotiation is, at bottom, a forced amendment of the psychological contract, executed without an opportunity to object. The written contract said you will use the new system. The unwritten contract said your expertise is valued here, your judgment matters, your standing is secure. AI adoption amends the second contract while pretending only to amend the first. The major gifts officer whose relationship instincts are displaced by donor scoring, the analyst whose synthesis arrives after the model’s summary, the editor whose judgment is now a review step on machine-generated copy: each of them is holding an amended contract they never signed, and each of them knows it, whether or not they can articulate it.

This is why so much of what we call resistance carries a moral tone that surprises leaders. People are not merely inconvenienced. They feel wronged. They feel betrayed. And here is the connection to the trust work from our midsummer sampler: every unilateral amendment of the psychological contract is a withdrawal from the trust account, and organizations spend that account freely while measuring it almost never. When the next initiative arrives, the account is empty, and leaders wonder why the announcement lands on folded arms.

The Three Clauses of the Renegotiation

If we take the negotiation frame seriously, then like any renegotiated contract it has clauses, and the organizations that are or will navigate this well are the ones that make each clause explicit rather than leaving it to be inferred from a slide deck.

The first clause concerns what is genuinely being surrendered. Leaders are understandably tempted to soften this in an attempt to gain acceptance. Nobody is being replaced, we say, the AI is just a tool, your expertise still matters. People know better, whether they are consciously recognizing this in the present or they are unconsciously accepting the present while worrying about the future and their role change, and the gap between what they are told and what they observe is measured in trust. The alternative is uncomfortable candor: this system will do a substantial part of what you currently do and doing that part of your work isn’t any longer how your value will be measured. That sentence costs a leader something to say. It also treats the listener as an adult party to a negotiation rather than an audience for a communications plan, and in my experience the candor is received better than the euphemism, because it matches what people have already concluded on their own. The leader is confirming, being honest, and vulnerable, which is what the individual is hoping for and what any rational person would expect.

The second clause concerns what is explicitly retained. Not retained by default, retained by declaration with all the necessary detail, at least as much as is known and can be communicated. Accountability is the clearest example. The renegotiation requires leaders to say, specifically and in writing where it counts: the model recommends, you decide, and when your judgment and the model disagree, here is who owns the call and here is how we will learn from the disagreement. Organizations that can’t complete that sentence have not adopted AI. They have abdicated to it as a result of market pressure, urgency for productivity gains, or the ambition to be seen as the innovator. The Confidence Transfer showed what follows: verification capacity atrophies, institutional expertise thins, and the organization discovers, usually during an incident, that it has automated its judgment without ever deciding to. Retention clauses are how that outcome is prevented, and they only work when they are explicit enough that a person can quote them clearly and in plain language.

The third clause concerns what is gained, and when, and how the gain becomes real rather than rhetorical. Ibarra’s research is practical here. New professional identity forms through doing, not through messaging, which means the compensation side of the renegotiation must take the form of actual work: the harder cases the machine can’t handle, the oversight and validation roles that require exactly the pattern recognition the veteran carries, the teaching of judgment to junior staff and to the systems themselves, the edge cases where twenty years of context is the difference between a defensible decision and a plausible-sounding error. If the elevated work doesn’t materialize on a visible timeline, people conclude the negotiation was conducted in bad faith and their deep, overthought suspicions have been validated.

What the Organization Loses When It Skips the Negotiation

Skipping the renegotiation carries organizational costs that compound quietly, and they connect directly to patterns this newsletter and the podcast have documented all year.

The first cost is the one The Confidence Transfer named: when the people who could verify the machine’s output no longer feel their verification is wanted, they stop offering it, and the organization loses its error-catching capacity precisely when it is deploying systems that fail in confident and plausible ways. The second cost is organizational memory. The veterans most affected by the write-down are the same people carrying the undocumented context, the institutional why behind the how, and Gallup’s 64 percent figure tells us where they are heading if the negotiation is botched. When they leave, the knowledge leaves, and as we explored in Organizational Memory Loss, almost no organization has any mechanism for noticing the departure until it needs what has already departed. The third cost is the trust account already discussed, which determines the exchange rate on every future change the organization attempts.

Add these together and the conclusion inverts the usual framing. The Identity Renegotiation is not a soft-side accommodation that slows down the real work of AI adoption. It’s the real work of AI adoption. The technology installation, the model refinements, the teaching of prompts are the easy parts, the technical challenges in Heifetz’s terms. The renegotiation is the adaptive challenge that determines whether the installation produces capability or theater.

What This Asks of Leaders

None of this yields a checklist, and I will not pretend it does. It yields a posture, and the posture begins with the diagnostic honesty most adoption programs are designed to avoid. The renegotiation lives in people’s values, habits, and self-concept. The Human Factor. You can’t project-manage your way through it. You can only lead people through it, which means creating the conditions Amy Edmondson’s work on psychological safety describes, where a twenty-year veteran can say I don’t know what I am now without that admission becoming evidence against them. It means sponsoring endings as deliberately as launches, giving the loss a venue instead of a euphemism. And it means accepting that the timeline for identity work is measured in months of doing, not weeks of messaging, and budgeting for that reality instead of being perpetually surprised by it.

So, the diagnostic questions, uncomfortable by design. Has anyone in your organization named, out loud, what your most experienced people are being asked to give up? Can each of them state which of their judgments still govern and which now belong to the system? Does the elevated work they were promised exist yet, and can you point to it, with names and dates rather than intentions? When your best people go quiet in the adoption meetings, are you reading compliance, or are you reading a negotiation you did not know you were losing? And if the renegotiation in your organization is currently being conducted in silence, who do you suppose is winning it?

The question for the rest of us is not whether our people will renegotiate who they are when the machines arrive. They already are, in every hallway conversation we are not part of. Lying in bed at night pondering their futures. Sitting in meetings drifting off in thought about what their role becomes.

The question we should all ask is whether we have the courage to join a negotiation that is happening with or without us, and to pay what candor costs before we collect what the technology promises.

Join the Conversation

How is the Identity Renegotiation being conducted in your organization, openly or in silence? I would love to hear your perspective. Share your thoughts with me on LinkedIn or subscribe to Ideas and Innovations on Substack for weekly insights on leadership, transformation, and the human side of organizational change.

The Truth About Transformation

If this issue resonated with you, my book The Truth About Transformation explores why most transformation efforts fail and what leaders can do differently. Available on Amazon.

Research Sources

1. Petriglieri, J. L. (2011). Under Threat: Responses to and the Consequences of Threats to Individuals’ Identities. Academy of Management Review, 36(4).
2. Ibarra, H. (2003). Working Identity: Unconventional Strategies for Reinventing Your Career. Harvard Business School Press.
3. Bridges, W. (1991). Managing Transitions: Making the Most of Change. Da Capo Press.
4. Rousseau, D. M. (1995). Psychological Contracts in Organizations: Understanding Written and Unwritten Agreements. Sage Publications.
5. Robinson, S. L., & Rousseau, D. M. (1994). Violating the Psychological Contract: Not the Exception but the Norm. Journal of Organizational Behavior, 15(3).
6. Heifetz, R. (1994). Leadership Without Easy Answers. Harvard University Press.
7. Edmondson, A. (1999). Psychological Safety and Learning Behavior in Work Teams. Administrative Science Quarterly, 44(2).
8. Gallup. (2024). Disruptive Change Is Hitting Leaders and Managers Hardest.
9. Novak, K. (2025). The Truth About Transformation (Revised and Expanded). 2040 Digital.