The Performance Review That Lost Its Empathy
The email looked routine: “Your 2025 review is ready.” The feedback felt less so. There was a neat paragraph about “cross-functional impact,” a sentence on “alignment with objectives,” and a line that somehow called my “communication clarity excellent.”
Only one problem. My manager had never used the word excellent in two years. That’s when it clicked: the system had written my review.
The illusion of objectivity
Across companies, performance management is quietly being rewritten by AI. Tools trained on competency parameters and sentiment models now write evaluation summaries in minutes. Managers simply approve and edit.
The promise is efficiency. Faster reviews, fewer human biases, standardized language. But the cost is subtle and enormous. The loss of emotional context.
AI can evaluate output. It can’t interpret effort, intent, or struggle. The human texture that defines a year of work gets lost. When algorithms flatten performance into sanitized adjectives, they erase the nuance that builds trust.
For employees, that trust gap feels like betrayal. The review no longer sounds like a conversation. It sounds like a report.
When empathy disappears from feedback
In a recent survey by Mercer, 61% of professionals said they would value a manager’s feedback more than an AI generated one, even if the latter was more detailed. The reason isn’t accuracy. It’s empathy.
What employees value more:
Human feedback (even if brief): 61%
AI feedback (even if detailed): 39%
Empathy lives in how feedback is given, not just what is said. A sentence like “You’ve grown a lot in navigating difficult clients” carries emotional memory. Shared calls, late nights, small wins. AI can replicate the phrase but not the experience behind it.
When leaders outsource feedback to automation, they also outsource the relationship. Over time, this changes what “manager” even means. The role shifts from coach to approver, from human mirror to algorithmic editor.
The convenience trap
The corporate argument is predictable: managers are overworked, employees want faster cycles, data beats subjectivity. But speed can’t replace meaning.
AI driven systems offer linguistic fairness. Everyone gets reviewed in the same tone. But not relational fairness. The quiet, personal calibrations never survive templated writing. “She struggled early but recovered fast.” “He took ownership after the setback.” These observations rarely make it through automation.
And once convenience becomes habit, reflection disappears. Reviews turn into compliance rituals, not growth moments.
Tools like Microsoft Viva and Workday are now piloting continuous evaluation features. AI tracking communication patterns, tone, and collaboration to generate real time performance nudges. The data may be right. But being right isn’t the same as being fair.
When algorithms start defining contribution, employees begin performing for the metric rather than the mission. A model can flag low participation in meetings. It can’t know your mother was hospitalized that month. It can summarize your achievements. It can’t sense your exhaustion or your resilience.
What’s being reclaimed
The companies gaining ground aren’t automating reviews. They’re using AI to surface insights while reclaiming the human ritual of reflection. Data measures performance. Conversation restores connection. The best performance systems will use algorithms to identify blind spots and managers to frame growth. They’ll let AI write drafts and humans write meaning. The truth is simple: employees don’t remember their ratings. They remember the moment their manager saw them for who they really were.
That’s something no algorithm, not even the most empathetic one, can replicate.
