When AI Replaces EQ: The Disintegration of Trust

In August 2025, researchers from USC and the University of Florida published a study of 1,100 professionals in the International Journal of Business Communication that confirmed what most people in team-based work environments had already felt but not formally named. Only 40% to 52% of employees considered supervisors who relied heavily on AI for workplace communication to be sincere, compared to 83% for those using minimal AI assistance. The trust gap between a manager who leans on AI and one who does not is not a nuance. It forms quietly, without a confrontation, every time a manager outsources the relational work of leadership to a language model. And it has an operational cost that neither the research nor most leadership development conversations have gotten around to naming directly.

The research captures the symptom. The more interesting problem is what happens to a manager's emotional intelligence development when AI becomes a standing substitute for the communication they find hardest. EQ is not a fixed trait. Six Seconds, using a validated 77-item assessment administered to 28,000 adults across 166 countries, documented a 5.79% global decline in EQ scores between 2019 and 2024, with statistically significant decreases across all eight measured competencies. The declines were sharpest in the skills most directly linked to motivation, optimism, and goal orientation. TalentSmart's research found that emotional intelligence explains 58% of performance across all job types, and a separate body of research ties approximately 77% of a leader's effectiveness to EQ skills specifically. The point is not that AI is causing this decline; the causal relationship has not been established and the timeline predates widespread AI adoption. The point is that a broad, measurable erosion in EQ was already underway before organizations handed their managers a tool that can produce polished, professional-sounding communication without requiring any of the emotional labor that builds the skill in the first place.

The pattern it produces is recognizable to anyone who has worked in a high-volume, high-friction operational environment. A manager who struggles with directness, empathy, or conflict will not stop struggling because their written communication now sounds measured and collegial. What changes is the visibility of the gap. Direct reports begin to notice a version mismatch: the communication they receive in writing reflects a level of emotional regulation and interpersonal consideration that does not show up in any live interaction. That discrepancy does not read as growth. It reads as inauthenticity.

In fractional and interim leadership roles specifically, the risk is compressed. A fractional or interim leader has a shortened window to establish credibility with a team that is already in some state of operational flux. Communication is the primary vehicle for that trust-building in the early weeks. When the written communication is polished and the in-person communication is dissonant from it, the team does not give the benefit of the doubt. They calibrate to what they observe.

The tells are consistent enough to be field-observable: an email opening with "I hope this message finds you well" from a manager who has never once written anything that warm in a Slack message; feedback framed in near-identical hedged constructions across multiple recipients; "I want to acknowledge your concerns and I appreciate you raising this" from someone whose verbal response to the same concern in a team meeting was dismissive; sentences that are perfectly parallel in structure and length in a way that human writing almost never sustains across a full paragraph; and the phrase "moving forward" appearing in the third sentence of every difficult email, regardless of context, because it is in every AI template for managing conflict. The tell is not the individual phrase. It is the voice discontinuity between the written version of this person and every other version you encounter.

What makes this consequential for operational performance, specifically in environments like healthcare operations and high-volume billing teams, is that trust between a manager and their team is not separate from the quality of the work. The Leadership IQ poll of 27,000 workers found only 23% believed their manager responded constructively when concerns were raised. That number was already low before AI gave low-constructiveness managers a tool to produce constructive-sounding responses. The gap between a response that sounds right and a manager who actually follows through creates the kind of disengagement that shows up in turnover, in quiet abdication of accountability at the work level, and in teams that stop raising the problems worth solving.

The argument is not against AI in professional communication. Using AI to edit grammar, improve clarity on routine informational messages, or restructure a complex announcement is legitimate use. The argument is against using it as a prosthetic for emotional competence a manager has not developed, because the prosthetic does not build the underlying capability and the team will eventually see through it. Leaders who want teams that perform because they trust the leadership have to do the difficult work in the difficult communication. That is where the skill is built. Outsourcing it is not efficient. It is a loan on credibility with compounding interest, and eventually the team calls it.