Upgrade.Digital
PDF herunterladen
Impulse

Translating When the Machine Becomes a Co-Translator

How the communicative repertoire of leaders is changing; and why proximity and presence, and the trust they carry, are becoming the real strength in the age of AI.

Von · · 15 Minuten

Größere Schrift, ruhigere Zeilen

A note on where this comes from: this text rests on many years of trusted work with leaders in large organizations, on collegial exchange with colleagues, and on current empirical research. It makes no theoretical claim; it condenses recurring observations from practice and sets them alongside the state of the research.

The Daily Craft

The communicative task of leadership has always been to translate reality. With AI, a second translator steps onto the stage: faster, smoother, clearer than any human; and carrying no responsibility. This changes the tempo of communication, and above all it changes the role of the leader.

Leadership is to a large extent communication. Not the ceremonial speech, the “state of the union,” which has its own function; rather the everyday act of translating. A directive from above is made intelligible for the team; a worry from below is made connectable for the level above; a number from controlling becomes an action a human being can carry out on a Monday morning. This translating is not a side activity. It is the substance leadership is made of.

In complex organizations this craft has always been demanding. A leader rarely speaks to only one audience: upward, downward, sideways; and across, along processes, value streams, and roles in the operating model. Each direction has its own language, its own interest, its own tempo. Anyone who wants to be heard in all of them has to hold the same reality in several versions without betraying it.

I. Leadership Translates: The Communicative Starting Point

Let us begin with a distinction that tends to blur in practice. When a message travels through an organization and changes its content on the way, at least three quite different effects are at work.

The first is the telephone game. A message exists, and it degrades in transit. Schulz von Thun described the mechanics early on: what is said is not what is heard; what is heard is not what is understood; what is understood is not what is agreed to. Across several levels these fractures accumulate, in both directions. Sometimes the message softens on its way down until urgency and substance evaporate. Sometimes the opposite happens: a requirement set with composure swells as it descends and triggers an alarm no one intended.

The second is faithful translation. Here too there is an original; but the leader deliberately re-encodes it so that it lands with the recipient. This is not distortion, it is competence. Ancona calls it relating and sensemaking: being visible, communicating constantly, building a shared map of the situation together, and saying openly what you know and what you do not. Good translation carries the relevance of a matter into the language of the person meant to hear it. Daily bread.

The third is the construction of reality; here no clean original is left. A leader holds several truths at once. “We have to become more innovative and increase cost efficiency” is one reading; “we have been running on fumes for a long time” is the other. Both describe the same organization. Experienced leaders hold them side by side without getting dizzy; virtuoso jugglers of competing logics of action.

These three readings do not exclude one another; in practice they lie on top of each other. But they make different demands: the first on care, the second on craft, the third on judgment. Whoever knows only the telephone game reads every deviation as a defect; whoever knows only construction treats every message as arbitrarily interchangeable. The art lies in between.

And it runs in all directions at once; that is the 360-degree character of the repertoire. Downward it is about movement: the team should act, not merely be informed. Upward it is about legitimacy; sideways, about interpretive authority. The board has barely adjourned before the levels are calling each other, and the message is not passed on but interpreted, growing denser with every loop, until the grapevine overtakes the memo. In a matrix a fourth direction is added, running along processes and cross-cutting roles, where the interest of the group meets the interest of the unit. Such conversations are negotiations at heart; and a large part of leadership communication is exactly that: negotiation.

That this translation work matters is not a consultant’s claim; it has been measured. The employee communication study by Staffbase and YouGov (2025, including a DACH sample from German-speaking Europe) shows that the direct manager is the most trustworthy source of information for employees, and that people who experience their manager’s communication as very clear are on average three times more satisfied in their jobs.

Clarity is not a question of style. It is a leadership instrument with measurable effect.

II. Appropriately Complex: Between Oversimplification and Noise

A distinction from systemic organizational development is worth recalling: complexity and complicatedness are not the same thing. Complexity is a fact of the situation: many elements, dense interactions, nonlinearity, unpredictability; it has to be matched. Complicatedness is homemade: the sediment of many individually sensible decisions, rules, committees, reports; it has to be reduced.

The same holds for communication. Ashby’s law of requisite variety states that one system can steer another only if it is at least as varied as the system it steers. Anyone who wants to steer a complex situation therefore needs a correspondingly complex repertoire. Gottfredson et al. confirm this empirically in The Leadership Quarterly (2025): complex environments require complex leaders; inner complexity has to match outer complexity.

For communication this means it must be able to carry ambivalence, and at times ambiguity, in order to do justice to the matter at hand. Two opposite errors follow, with a narrow ridge between them.

The first error is under-complexity: a false simplicity that contradicts the reality people experience. It costs trust, because recipients feel the gap between the polished message and the rough situation immediately. The second is over-complication: communication that adds nothing to the situation except noise, caveats, footnotes, hedges. It suffocates the very movement it is meant to set off.

An everyday example shows both errors at once. Almost every organization has to save and invest at the same time. What arrives with the employee is: they are cutting on my side while money is being spent over there. The experience of contradiction is real; ignoring it would be under-complex, drowning it in caveats would be noise. The task is to dissolve the contradiction: to show that an organization, like a private household, has to save and invest at once; that the two do not refute each other but depend on each other. In private life we tolerate this simultaneity effortlessly. In corporate reality the logic sometimes deserts us. Bringing it back is communicative work.

A trap is waiting here. The leader could side with the team: they are cutting too hard around here. That earns short-term sympathy, quick Robin Hood points. But it does not make the message any clearer; it puts the message at odds with what the organization is visibly doing.

Under-complex communication does not reduce complicatedness. It produces it. Today’s supposed simplification is tomorrow’s noise.

III. Held Tension: Both/And as a Craft

Once ambivalence becomes a requirement, a misunderstanding is close at hand: that both/and means the middle. It means the opposite.

Systemic organizational development, for instance in the work of Rudolf Wimmer, describes the capacity not to resolve competing demands but to hold them in tension. Paradox research, prominently in Smith and Lewis, calls it both/and thinking: contradictions such as stability and change, center and unit, saving and investing are not problems to be solved by taking sides. They are permanent tensions to be managed.

For practice: held tension is not the slider. The slider averages: save a little, invest a little, and otherwise everything continues; it downgrades the reorganization into a minor adjustment and drains both messages of force. Held tension does the opposite: it sharpens both poles. Not “we will save a little,” but “here we are cutting hard and deliberately, and in concrete terms that means these priorities”; and in the same breath, “there we are investing without compromise, because our future depends on it.”

Balancing is the trap. Sharpening both ends is the craft.

This allows us to sharpen a term one hears often: calibrated vagueness. The word misleads; it sounds like a smokescreen. The opposite is meant: not distortion, but dosage in sequence and emphasis. What is said first, what becomes the main subject, what becomes the secondary condition? Whether the cut appears as the price of the future, or the future as a gloss on the cut, is decided by the order, not by the adjective. Calibrated weighting is therefore the more accurate term; it separates the craft from the smoke screen.

IV. Two Stages: The Bound Front Stage and Movement on the Inside

This calibration happens on different stages. Goffman’s distinction still describes organizations precisely: a front stage on which a coherent performance is given; a back stage where the actual work happens, the improvising, the failing, the fixing. The front stage, too, can be played well or badly.

On the front stage, facing the supervisory board, the capital market, and the public, the message is tightly bound. Not out of cowardice, out of necessity: ongoing negotiations, legal risk, disclosure obligations, the protection of the people affected, what can be said toward a competitive market. Balance here is not evasion, it is a constraint of the role.

Internally, however, facing the team, the unit, the business line, all of whom are meant to act, that same balance does not pay off. Telling a workforce that feels the reorganization in its own body that things will be “a little different, everything continues” produces no movement – it produces the impression that reality is being denied.

Internally, communication is not an expression. It is a lever for producing movement.

The conflict is not resolved by choosing a stage, but through a second capability. The reflective leader does two things. She legitimizes the irritation that her clearer internal communication provokes upward; and she makes transparent to her team that different places call for different points of emphasis. This disclosure is itself a communicative achievement, and it has a limit: if the versions contradict each other too strongly, one discredits the other.

The real act of translation is not producing two versions; almost any experienced leader can do that. It is disclosing that the versions exist without devaluing either of them. Translating then means not concealing, but making the translation itself visible.

V. The Machine Is Listening: How AI Shifts the Communicative Reality

This whole two-stage craft rests on one precondition: the back stage was not automatically visible from the front stage. That is exactly the wall artificial intelligence is taking down.

The briefing does not lose its function, but it loses its exclusivity. Anyone who receives an official version can build their own AI-assisted analysis within minutes, one that holds its own in depth and breadth. What was backstage yesterday now sits on call, right next to the official account. The bound message no longer sits behind a wall, it sits next to a reconstructable truth.

For now, the AI systems provided in-house lag behind; many models are not connected deeply enough to internal databases and bodies of knowledge. Gartner (2025) reports that only a portion of organizations reliably move AI beyond the pilot stage into production, and that trust decides between success and failure; according to the CIO survey (2026), only around 17 percent have AI agents in use, while more than 60 percent plan to within two years. In reorganizations, downsizing programs, new appointments, and strategic shifts, the leader still has a hand on the valve. But the shift has begun, and faster than we assume.

Second, what changes is less the information than its value. If everyone can call up the state of affairs, the leader loses a classic source of power: not what she knows, but when, to whom, and how she releases it. The lever of sequence and timing erodes; interpretation takes its place. Information is not scarce – its interpretation is. Framing, the narrative share of leadership, gains weight.

AI intervenes in leadership communication on three levels.

Level one: AI as a source of information and narrative for employees

Employees increasingly get their information through AI, and not only facts but narratives. The systems deliver explanation, context, and story along with the answer. The Staffbase study notes that AI now answers employee questions directly, and that the broadcast model, the same message to everyone, is eroding as a result. Answers are expected immediately and tailored. The already large share of the working day spent on pure information consumption (Gartner) is being shifted by AI to exactly this point.

For leadership this means the advantage of informing first and exclusively is melting away; at times the leader is catching up with what the team already believes it knows.

Level two: AI as sparring partner and turbo for the leader’s own repertoire

From the leader’s perspective, AI is at the same time a tool. It can prepare messages, test them for clarity, play through perspectives, sort out conflicting interests, and draft formulations for different audiences. In daily sparring, one’s own repertoire can be tuned enormously: This is lived reality already, not a promise about the future. The flip side is delegated, synthetic communication: where preparation migrates to the machine, the relationship thins out; the direct interaction in which trust is built gives way to processed messages.

Level three: AI as communicative diagnostics and organization-wide alignment

The level perhaps most underestimated is diagnosis. AI-supported dialogue assistants can take internal temperature readings: how is a topic actually being received in the house? Until now leadership has had to rely on attribution and gut feeling, on what it observes in the hallway, or on surveys with rigid questions whose dialogic share often approaches zero, even when some of them carry “dialogue” in the name.

In the future it will be possible to determine more precisely what was heard and what was understood, and the patterns and sentiments behind it will become visible faster. That opens up the possibility of bringing an organization’s often contradictory constructions of reality into a workable consent: becoming clearer without wiping the complexity away. Used well, this level pays into trust.

So much for the opportunities.

The most comprehensive independent study to date, “Trust, Attitudes and Use of AI” from the University of Melbourne with KPMG (2025, more than 48,000 respondents in 47 countries), shows a double picture. AI is widely used and experienced as helpful; at the same time, 57 percent rely on its results without checking them, 59 percent have made mistakes as a result, and almost half use AI in ways that break the rules. On first contact, clarity beats truth.

The machine earns trust not because it is more correct, but because it sounds clearer.

And there is a relational flip side. Peer-reviewed studies on employee-AI collaboration (2025 among others) show that it can increase loneliness and emotional exhaustion: the machine takes over tasks that used to require exchange, and the significance of the relationship with colleagues declines. Emotional support from the leader reduces loneliness. The finding leads straight back into leadership communication.

Where the machine costs proximity and presence, the leader’s presence, and the trust they carry, become more valuable, not redundant.

VI. What Remains for Leadership: The Repertoire Under AI Conditions

Let us draw the lines together. When information becomes cheap and clarity becomes mechanical, the value of leadership shifts. Not toward more clarity; the machine supplies that effortlessly, often more smoothly than any human. It shifts toward what the smooth machine precisely does not supply.

Framing, the cognitive side of translation, AI can help with: suggestions, variants, sharper versions. Proximity and presence, and the responsibility a person assumes, are the relational side; that side it cannot supply. Trust does not arise because someone finds the best formulation. It arises because someone stands behind it.

From this follows a shift that is already becoming visible. Leaders who create proximity, presence and trust with virtuosity will hold their ground; their repertoire is amplified by AI, not replaced. Those who are weaker enter a competition they will struggle to win: where communication has so far been deliberately diplomatic, balanced, and blurred, it will now be measured against the clarity of the AI answer. The perpetually slightly vague language of some leadership levels holds trust less well when a machine next door answers clearly. Not because the machine is right, but because it sounds as though it were.

And how do you recognize a leader whose communicative repertoire is equal to the age of AI? Without any claim to completeness:

  • She names contradictions before she classifies them; she does not conceal the price of her own priorities.
  • She sharpens both poles of a tension instead of averaging toward the middle.
  • She discloses that different stages require different versions, and she reconciles those versions instead of sealing them off from each other.
  • She uses AI for preparation and for diagnosis, not as a substitute for the direct relationship; proximity and presence remain the core task, and there is now more time for them.
  • She checks what the machine delivers instead of trusting it because it sounds clear; and she makes that checking visible to her team.

Translating While the Machine Listens In

Leadership was always translation. What is new is the co-translator in the room, who formulates more clearly than any human and carries no responsibility. That devalues the part of the old repertoire which rested on an information advantage; and it raises the part which rests on held tension, on interpretation someone answers for, and on the proximity and presence that produce trust.

The translation trap of the AI age is therefore not explaining too little. It is relying on the clarity the machine supplies so easily and losing the ambivalences that make up reality in the process. The strength of leadership emerges once you expose it: holding several truths at once, naming the price of a decision, and standing behind an outcome you do not fully control. Which is exactly why it becomes more valuable.

Sources

  • Deborah Ancona et al., the 4-CAP/4-CAP+ model of leadership capabilities (sensemaking, relating, visioning, inventing; extended by “building credibility”), MIT Sloan School of Management.
  • Ryan Gottfredson et al., “Navigating complex environments requires complex leaders,” The Leadership Quarterly (2025).
  • Staffbase & YouGov, “International Employee Communication Impact Study 2025” (including a DACH sample from German-speaking Europe).
  • Nicole Gillespie, Steve Lockey et al., “Trust, Attitudes and Use of AI: A Global Study 2025,” University of Melbourne & KPMG (more than 48,000 respondents, 47 countries).
  • Gartner, findings on AI maturity (Survey 2025) and CIO and Technology Executive Survey (2026).
  • “Effects of Employee-AI Collaboration on Counterproductive Work Behaviors: Leader Emotional Support as a Moderator,” peer-reviewed study on loneliness and emotional exhaustion (MDPI, 2025).
Teilen LinkedIn E-Mail