My company is considering using AI detectors to review the self-assessments submitted by employees for promotions. I think this is a terrible idea because our internal templates are so rigid that everyone’s writing looks the same. Won't this lead to unfair results?
3 answers
Rigid templates are a "green flag" for most AI detectors, making them a nightmare for internal promotion cycles. If your company requires everyone to use the same "Situation-Task-Action-Result" (STAR) format, you are essentially forcing human employees to write in the same structured way that LLMs are trained to output. This will lead to a high volume of false positives. A much better approach for the PMO would be to evaluate employees based on their delivered KPIs and peer reviews rather than the linguistic "randomness" of their self-assessment documentation.
How are managers supposed to handle it if a top performer is flagged by the software as being "artificial"?
Templates are for consistency. Using a detector to penalize that consistency is completely counter-productive.
I agree with Anthony; the goal of a template is to make things easier to read, not to provide "creative" fodder for a detection algorithm.
That is a management nightmare, Steven. It creates a "guilty until proven innocent" culture. If a manager trusts their employee but the AI detectors say otherwise, it undermines the manager's authority and destroys team trust. Companies need to realize that these tools are probabilistic, not deterministic. They should never be used as the "sole judge" for something as important as a person's career progression or professional reputation.