Payscale tells LDS why new Software Engineer V hires out-earn the veterans by 19%

The Software Engineer V result is an exception inside a family that otherwise rewards staying. Thomas framed it as competition concentrated at a single level.

"Software Engineer V is the highest flight-risk role in software development," she told Lets Data Science. "That's a 19% gap because organizations are competing aggressively for senior engineers who can build and scale complex systems. AI is raising the ceiling on what those systems need to do, so demand is intensifying even as some routine coding faces automation pressure."

Data work behaves differently. "Data science tells the opposite story," Thomas said, with the report showing tenured data scientists ahead by 21% at the job-family level. "Same disruption, but different scarcity profile."

The most counterintuitive answer concerns where the money has actually moved, and it is not toward AI skills in engineering.

"The biggest premiums aren't always where you would expect," Thomas said. "In core tech, AI fluency is already assumed. Everyone's expected to have it, so it doesn't move the needle on pay."

Instead the premiums surface in places that are not thought of as AI roles at all. Thomas pointed to Marketing Analyst III, IT Product Manager IV and Senior Compliance Specialist, where Payscale sees "8-12% new hire premiums in functions where AI capability is still genuinely scarce."

There is a seniority gradient underneath it. Entry-level new hires earn roughly 5% less than tenured peers, while at the C-suite new hires command a 3.3% premium. "Compensation is shifting toward rewarding the ability to apply AI in ways that create business value instead of technical fluency alone," Thomas said.

The report is framed against fears that AI is compressing pay. Thomas's answer splits the losers into two very different groups.

The first is the expected one. "Routine, repeatable work is getting easier to automate. Pay and opportunity at the entry level are compressing."

The second is not. "There's a second group losing too, tenured employees in roles like Software Engineer V, where the market has moved faster than internal pay structures have kept up," she said. "Their skills aren't obsolete. They're exactly who companies are fighting to hire externally. The internal comp systems just haven't caught up."

What inflates pay, in her account, is scarcity rather than AI adjacency. She offered Training Coordinator as the illustration, where new hires earn 10% more than tenured staff. "That's not an AI role. It's a reflection of how hard it is to find people who can lead AI adoption and workforce enablement at scale."

Asked whether it still pays to build tenure or to switch and specialize, Thomas refused the general answer.

"It depends on where you sit relative to supply." Tenure still wins across most of the labor market, with employees in tenure-advantaged roles earning 6.1% more on average than new hires. "Staying and letting expertise compound is still rational for a lot of careers, especially early on."

The exception is genuine scarcity, particularly in places employers did not expect to need the capability. Her test is refreshingly concrete: "This is less about learning AI and more about understanding what's actually valuable about what you do and whether the person next to you has the same skills. If they do, tenure and institutional knowledge are your best lever. If they don't, you have room to move."

Thomas was blunt that new hires out-earning incumbents is an internal equity problem that will not stay hidden.

"When two people are doing comparable work and one earns meaningfully more just because of when they were hired, your existing team will notice. Especially in an era of pay transparency where salary ranges are visible in job postings."

Her assessment of the current corporate response is that it is stalling. Most companies are reaching for hiring premiums, some upskilling, and a reluctance to make permanent base-pay adjustments for skills they suspect may be commoditized later. What she recommends instead is visibility into where these skills are used and paid for, internally and externally, plus governance that sets in advance when a premium is justified, "so they're not making reactive decisions."

Payscale drew the report from its proprietary Peer dataset of HRIS records covering 10.2 million incumbents, 4,500 participating organizations and 36,000 jobs, effective May 2026. New hires had been in role no more than a year, tenured employees at least 13 months, and published national medians required at least 40 incumbents plus confidentiality safeguards.

One caveat worth keeping. Payscale links the changing gaps to AI-driven skill shifts and slow-moving internal pay structures, and that is the company's interpretation. Cross-sectional pay comparisons cannot on their own establish that AI caused any particular gap. The practical signal for practitioners is narrower: benchmark at your exact role and level, rather than treating an industry-wide AI salary narrative as a substitute for current data.