How Companies Are Using AI to Set Salaries

Artificial intelligence is already changing one of the most important parts of a career: how much money people make. Compensation surveys suggest the average pay increase next year will be about 3.5%, but the spread of raises around that midpoint will be wider than usual, according to a Wall Street Journal report. Where an employee lands may come down to what a machine-learning model thinks they are worth.

Salary benchmarking as a service

More firms are using AI tools to comb job boards for competitors' salary ranges, helped by the spread of pay-transparency laws. Third-party companies now offer AI salary benchmarking as a service. Stello AI analyzes public compensation information from job listings and forums such as Glassdoor, plus data purchased from payroll companies, to assess the pay of every employee at a new client.

"We tell you exactly how much to pay your employees," says Chief Executive Amee Parekh. One group gets special attention: "Here are your high performers who are underpaid. If you have any budget for raises, these are the first ones you should fix." People who could earn more elsewhere are flight risks.

Overpaid workers and modest raises

Stello's analyses also identify the opposite category: workers who are overpaid, often people who job-hopped at an opportune time and scored packages more generous than today's tepid labor market would bear. They are unlikely to be fired for making too much, Parekh says, but they should brace for a modest raise, if they get one at all.

Good managers will explain small raises to well-paid workers, says Tauseef Rahman, U.S. workforce reward solutions leader at Marsh, formerly Mercer. "They'll say, 'You're already at the top of the range, so this year you will not be getting as much. We still value you, and we should be talking about your promotion path instead.'" A Marsh survey of more than 1,000 U.S. employers found the average planned merit increase for 2027 is 3.2% of a total raise budget of 3.5%. The Conference Board found nearly the same, with firms rewarding AI savvy and hard-to-replace skills.

Rethinking the compensation model

AI is also disrupting the metrics that traditionally factor into pay decisions, such as billable hours in professional services: should paychecks shrink if AI reduces the time needed for certain tasks? Lesley Uren, chief executive of Korn Ferry Consulting, says clients are wrestling with such questions. "You have to really rethink what good looks like, what success looks like," she says.

One vision is to tie pay more closely to impact and outcome and de-emphasize activity measures such as hours worked. That is complicated because impact can be intangible: AI could make an employee's value more subjective even as it tries to objectively set the going rate for their services.