Is AI Mucking Up Performance Reviews?

The employee thought she had put in a stellar year, exceeding all of her goals and volunteering for some key assignments. But her performance review scarcely mentioned any of this. She couldn’t understand how her manager could have gotten it so wrong.

The culprit may not have been a manager per se, but rather a manager’s overreliance on artificial intelligence. In the latest example of AI disrupting business, more and more firms are using the high-tech tool to process performance reviews and help them evaluate employees. It’s an effort to bring transparency and objectivity to a process everyone agrees is broken—only 2% of HR pros think performance reviews work, according to one recent poll. The problem, say experts, is that managers, who have been stretched too thin as firms lay off middle management, may not be confirming the tool’s accuracy or considering critical variables it has omitted. “I’m sure some leaders use the material AI prepares ‘as is’,” says Ilene Gochman, leader of global expert assessment and succession solutions at Korn Ferry.

Companies are already using AI tools to aggregate data and rate performance based on metrics such as sales, engagement, turnover risk, and pay. When they prepare for reviews, many managers and HR pros feed objectives and notes into AI to help them draft talking points on employee strengths, weaknesses, and needed development. The hope is that conscientious managers are using AI as a starting point and personalizing the text it generates to each employee’s circumstances, capabilities, potential, and more. That’s not always the case, though, says David Farris, global account leader and sector lead for professional services at Korn Ferry.

In large organizations with tens of thousands of employees, performance reviews are a massive undertaking. Farris worries that managers “are taxed for time and taking a cursory view of the data and making assumptions.” In a very basic example, AI may show that an employee didn’t meet sales goals, but may not know that they were pulled into a special assignment for three months. Or it may not be aware why a worker on maternity leave is missing meetings.

Bryan Ackermann, head of AI strategy and transformation at Korn Ferry, says companies are “losing the plot” regarding performance reviews: They’re not simply a rote management grading exercise, he says; their purpose is to help develop talent and promote growth, both for the person and organization. One of employees’ biggest complaints about performance reviews, for instance, is that managers’ ratings and feedback are arbitrary and unclear. Ackermann says AI can help with that by creating more transparency in rankings and customizing action plans for employees.

Ackermann describes a performance review as a “low-trust exercise screaming for more human interaction,” and worries that using AI in this context is a slippery slope. While AI can help save time and standardize data around objective metrics, it can’t make employees feel valued, determine an overall picture of their contributions, or inspire action. “I worry about any organization that believes it can take performance reviews from managers and give them to a bot,” he says.

He recommends that firms pair objective ratings from AI with developmental guidance from managers. He also advises companies to think about the main pain points for performance reviews and how AI can solve for them. On the technical side, organizations can ensure that the data AI uses is seamless and integrated so that insights can be shared across dashboards (for instance, so that AI can recognize, where noted, that someone missed their sales goals because they were working on a special project). "Performance reviews are really where that science of AI and the art of management merge," says Farris. 

Learn more about Korn Ferry’s Assessment and Succession capabilities.

Read more This Week in Leadership articles