Employers Turn to AI Benchmarking Tools to Decide Who's Underpaid or Overpaid, WSJ Finds
Companies are increasingly turning to AI benchmarking services that aggregate public job listings and payroll data to flag which employees are underpaid or overpaid, according to reporting from the Wall Street Journal's Callum Borchers. The tools feed results directly into pay-band decisions, a shift from the annual survey-based compensation reviews that have long been standard in corporate HR departments. The change is landing as pay for AI-related roles splits sharply, with an enterprise tier running $170,000 to $245,000 and a frontier-lab tier reaching $600,000 to more than $1 million.
The shift cuts both ways for workers. A separate preview of Payscale's AI Workforce Impact Report found 61% of organizations are rewriting job descriptions because of AI, but only 48% say their current market benchmarking actually reflects the skills those rewritten roles now require. At the same time, employees are increasingly walking into review conversations with their own AI-generated salary ranges pulled from tools like ChatGPT or Gemini, sometimes citing figures that don't account for a company's budget constraints or internal pay equity structures.
The result is a compensation system where both employers and employees are leaning on algorithmic outputs to argue their side of the same negotiation, often without clear visibility into how either tool arrived at its numbers or what data it drew from.
What supporters say:
Real-time AI benchmarking can catch pay gaps faster than annual surveys, which are often outdated by the time they're published.
Standardized data-driven pay bands can reduce the role of manager discretion, a factor that has historically contributed to unequal pay outcomes.
Faster benchmarking helps employers stay competitive in a labor market where AI-related roles are repricing quickly.
What critics say:
AI benchmarking tools often pull from inconsistent or outdated public data, producing numbers that don't reflect a specific company's budget or internal equity structure.
Workers have little visibility into how these systems flag them as under- or over-paid, and few companies offer a way to contest the outputs.
Using algorithmic benchmarks to identify "overpaid" employees risks becoming a tool for justifying pay cuts or targeted layoffs rather than fair-pay corrections.
What's your take?
Should employers use AI benchmarking tools to help set employee pay? Yes 
No 
Other 
Sources:
#ArtificialIntelligence #FutureOfWork #PayEquity #TechPolicy #Compensation
Now let's hear from you.
You get one Take and 3 ratings so use them well.
Strong arguments beat loud ones. See Moving The Needle (How to)
Don't forget to rate your own Take.
Please don't feed the trolls.
