AI is changing what companies need from their people. The question is how quickly that change is actually showing up in the workforce — and whether the way companies define, pay, and hire for jobs is keeping pace.
Oyster’s 2026 Global AI Talent Report looks beyond predictions about the future of AI to examine what is happening in jobs right now. We analyzed approximately 15,000 job descriptions for roles filled through our platform between 2022 and the first half of 2026, and paired that proprietary data with a third-party survey of more than 1,000 global HR and People leaders, along with interviews with experts working at the intersection of AI, talent, and global employment.
Together, the dataset reveals a significant shift already underway.
AI requirements have moved into established roles
In 2022, approximately 5% of roles filled through Oyster explicitly required AI skills. In the first half of 2026, that figure jumped to 15%. That is a tripling in four years.
What is especially significant is where those requirements are appearing. AI fluency is no longer confined to engineering or specialized AI positions. Engineering roles still lead at 23%, but 17% of marketing coordinator roles, 15% of recruiter roles, and 10% of operations coordinator roles filled through our platform now explicitly require AI skills.
As Jess Von Bank puts it in the report: “When we say ‘AI talent,’ we make the mistake of referring to technical roles. Code isn’t the bottleneck anymore—capability is. Every single function in the enterprise is becoming AI-native, whether we’ve redesigned the job descriptions or not.”
The data reveal a gap between expectations and workforce practices
The new expectations for AI skills are often not accompanied by changes in how jobs are defined, communicated, or compensated.
Ninety-five percent of surveyed HR and People leaders say employees at their company use AI tools daily in roles where AI is not part of the official job description. Among companies that added AI requirements to existing roles in the past year, 62% say those changes came with neither a new title nor a change in pay. And among companies extending AI requirements into non-technical roles, 46% say the new expectations were never formally communicated to employees.
The pattern also runs against seniority assumptions. AI skill requirements appear in 11–13% of individual contributor roles, compared with 8–9% of Director, VP and C-suite roles — concentrated where the day-to-day work happens, rather than where decisions about that work get made.
The shift in what companies ask of their people is not yet reflected in how those jobs are defined.
The AI talent search is becoming a global workforce story
At the same time, employers are making choices about how to close emerging AI skill gaps. More than half (53%) prefer to train existing employees when AI skills are missing. Roughly a third combine training with external hiring, while 14% primarily recruit externally.
Those choices do not just determine who companies hire. They are beginning to change where companies look. Companies prioritizing external AI hiring are nearly twice as likely to have recruited internationally because they could not find AI-ready candidates domestically: 34%, compared with 18% of companies prioritizing internal development.
And the markets where AI skills appear most often in filled roles are not always the ones that make a shortlist. In Oyster’s platform data, approximately 21% of roles filled in Switzerland and 20% in Poland require AI skills, followed by Portugal at 19%, Romania at 17%, and the Netherlands at 16%.
As Oyster CEO Hadi Moussa says in the report: “Companies can’t keep looking for the same skills in the same places and expect to stay ahead. The opportunity is to build AI fluency across your workforce—and open the door to talent you might never have considered before.”
What it means for people teams
The findings point to a practical starting place. Before opening a job requisition for AI skills, it is worth establishing what capability already exists inside the organization — because in most companies, some of it is already there and undocumented. From there, update job descriptions that have gone stale, decide deliberately whether AI fluency affects compensation, and tell people plainly what the job now requires.
The workforce is already changing. The question for people teams is whether job architecture, pay structures, and hiring criteria can catch up.