Where AI pay is heating up in 2026: Global shifts and emerging markets - HR Executive

A new global survey from WTW puts numbers behind why the market for AI talent is fragmenting, and a pay strategy built on last year’s data is already behind.

The firm’s 2026 Artificial Intelligence and Digital Talent Salary Survey, covering 10 countries, finds that median total compensation for mid-level machine learning roles in the U.S. now exceeds $170,000, compared to roughly $122,000 in Germany and just under $100,000 in the UK. Canada, once a close competitor, has slipped to fourth place and actually recorded declines in median pay for these roles.

Across all markets studied, total compensation for machine learning roles grew 6% on average, but base salaries moved just 2%. This factor may signal that incentives, not raises, are increasingly doing the heavy lifting in retention.

“AI pay is no longer just about where salaries are highest, but where momentum is building fastest and how employers are aligning pay and incentives to keep pace,” wrote Lesli Jennings, North America leader for work, rewards and careers at WTW, in a release.

“Employers that rely on last year’s assumptions risk falling behind, particularly as short and long-term incentives play a bigger role in fast growing markets.”

Mexico posted a 19% rise in base salaries and a 29% jump in total compensation for machine learning roles, the steepest growth in the study. Brazil also saw double-digit increases. WTW attributes emerging market momentum in part to expanding infrastructure investment and growing employer appetite for cost-effective AI talent.

Meanwhile, cloud computing pay is accelerating too. Median salaries for cloud engineering rose an average of 9% across the 10 countries, with total compensation up 12%. China and India drove much of that growth.

Read more: What Meta’s visa filings tell HR leaders about the real cost of AI talent

Global skills shortages

While WTW’s survey tracks where pay is breaking away, ManpowerGroup’s 2026 Talent Shortage Survey shows why pressure isn’t likely to ease anytime soon. In a poll of more than 39,000 employers across 41 countries, AI model and application development and AI literacy now rank as the hardest skills to find globally, overtaking traditional engineering and IT.

Nearly three-quarters of employers still report difficulty filling roles, despite a slight easing from last year. That scarcity is most acute in markets like Germany and the UK, where reported shortages exceed 70%, reinforcing WTW’s view that local conditions are driving today’s AI pay decisions.

What HR leaders should act on

Base pay alone is losing ground as a recruiting tool, according to the report. Nearly half of organizations now offer differentiated reward programs for digital talent, and long-term incentive vehicles like restricted stock units with regular vesting schedules are emerging as a key retention mechanism, particularly for AI roles.

Supply and demand for AI talent remain highly concentrated. The U.S., India and Germany lead demand for both AI engineers and machine learning engineers. India has the deepest supply, followed by the U.S. That concentration means local market conditions, not global benchmarks, should drive pay decisions.

And despite the attention AI commands, software engineers remain the most in-demand digital role globally, followed by application developers and data scientists. AI and machine learning engineers rank lower in current demand, though WTW expects that to change as adoption deepens.

“These patterns underline why a single global pay strategy rarely works,” Jennings said. “What is considered a hot role depends heavily on local supply, maturity of adoption and the mix of incentives on offer.”