How AI performance assessment works

AI performance assessment typically aggregates quantitative data — number of emails sent, response times, meeting attendance, sales figures, productivity metrics from software — and generates a score or ranking. More sophisticated systems analyse communication patterns, sentiment in written communications, and activity levels across multiple systems. The most advanced systems attempt to infer factors like engagement, collaboration quality, and leadership potential from observable proxy metrics.

The fundamental problem with AI performance assessment is that it measures what is measurable, not what is important. An employee who works fewer hours but produces higher quality strategic thinking, who takes time to mentor colleagues, or who builds relationships that generate long-term business value will often score lower than an employee who generates high volumes of easily measured activity. AI performance systems optimise for what they can see, not for actual performance.

Your right to an explanation

Under GDPR and UK GDPR, if you have been subject to an automated decision that has a significant effect on you, you have the right to obtain "meaningful information about the logic involved" in the decision. For an AI performance system, this means your employer must be able to explain to you — in terms you can understand — what factors contributed to your performance score, how those factors were weighted, and how the score was calculated. An explanation that consists entirely of references to "our proprietary AI model" or "our performance management system" does not satisfy this requirement.

In practice, many organisations cannot provide adequate explanations of their AI performance systems because the systems were procured from vendors whose methodology is not transparent even to the employer. If your employer cannot explain how your performance score was generated, they may be in breach of their GDPR obligations — and you may have grounds to challenge not just the specific score but the use of the system.