With the help of AI, some companies are bringing more structure and strategy to how they compensate employees.
A 2025 survey from Korn Ferry found that roughly a quarter of the 5,717 companies they surveyed are using AI to help determine compensation. Though just 22% of surveyed companies said they are using AI for external-pay benchmarking, 63% say they are considering using it.
Ruth Thomas, the chief compensation strategist at the software company Payscale, said that when companies use AI to analyze large amounts of compensation-related data, it can help promote pay transparency and provide HR teams with an additional tool for understanding new and changing job markets.
At the same time, human resources workers should still audit and monitor AI-driven data tools to keep confidential information safe and avoid data errors that can lead to pay bias, said Gord Frost, the global rewards solution leader at global consulting firm Mercer.
AI can assist with filling compensation gaps
At Payscale, the company uses a combination of AI modeling and HR-contributed salary data to help its customers price jobs, Thomas said.
The tool that does this, Payscale Verse, is especially useful when data is limited because a job is new to an industry or has rare requirements for its location, company size, education, or experience level, Thomas said.
“Sometimes our jobs are really niche, and we have a hard time finding matches for them,” said Kristen Damerow, an HR analyst at SmithGroup, an architectural firm that uses Payscale’s services.
There’s also Payscale Peer, a dataset built from salary information that includes compensation data from more than 5,400 organizations, Thomas said. This data is pulled daily from human resource information systems, or software platforms that help companies manage their operations. It’s 100% employer-reported, which is different from job-posting data that other compensation vendors use, she added.
Peer can show Payscale users what the market is paying for certain roles now, as opposed to purchasing a salary survey from a survey provider, where the information can be outdated. Thomas said if a compensation manager is trying to set a salary for a brand-new job but doesn’t have much data about that role, Payscale Verse uses Payscale Peer’s data and AI modeling to find similar roles in different locations. Then, the AI algorithm takes those differentials and suggests a market price to match the new job.
By using AI, compensation managers can more quickly and efficiently see how pay compares by location, industry, and size. For instance, if a company isn’t sure how much to pay a culture experience specialist in the hospitality industry, Payscale can fill in the gaps by pulling data from a similar field, like travel and tourism.
Thomas told BI that after Payscale Verse recommends a suggested salary match, the company decides whether to accept it. She said firms have accepted an estimated 88% of Payscale’s AI-recommended matches, compared with the 12% that were accepted previously when Payscale started using the technology.
AI for automating HR tasks
Frost said AI also has the potential to automate HR professionals’ routine and repetitive tasks, such as submitting salary data for annual compensation surveys and retrieving benchmark data to compare pay across companies and roles.
For instance, Frost said that by having more timely and robust access to external market data, rewards teams — which are responsible for designing, implementing, and managing programs that recognize employees for their contributions — would be able to react quickly to changes in the talent market by making salary adjustments in real time.
Frost said HR professionals can better understand which elements of their total rewards programs have the most impact on employee retention and performance, and then focus on investing in programs that resonate the most with different employee groups.
He added that an HR team could also use AI to generate personalized talking points for managers to explain pay decisions and the company’s pay programs more consistently with employees across the organization.
Weighing the risks
Though AI can help determine compensation, human oversight still matters.
Thomas said that before data is added to Payscale’s database, it’s validated using a series of automated outlier detection steps. She added that human reviewers also routinely audit data.
“All of our tools are built with pay transparency in mind,” Thomas said. “Payscale works hard to help organizations understand where the data comes from and how we use data to arrive at salary information.”
The rise of AI in compensation has also led to new questions about how vendors are building and using the tools. Thomas said that HR professionals should work with their vendors to ensure that they understand how each vendor is using AI in its compensation management solutions.
“The onus is on the employer to make sure that they are aware of any potential bias in the vendor solutions,” Thomas said. For example, if an AI model were trained on historical salary data that shows men earning more than women for the same job, it could unintentionally recommend lower salaries for female employees.
Frost said that HR professionals need to be aware of the risks of using AI, especially when it comes to protecting confidentiality while comparing employee data or using salary analytics tools.
“These are the kinds of responsibilities that total-rewards teams take seriously, and while AI is a powerful tool that can help with the process, the importance of the human element cannot be overlooked,” Frost said.
Read the full article here
