Not all AI skills are earning the same premium right now, new data shows
The conversation around artificial intelligence at work has shifted. Companies now expect AI use to translate into measurable business results, a higher bar than simply confirming the tools are being used.
New research from the Upwork Research Institute points to a specific kind of worker emerging to meet that demand, the AI orchestrator.
Jennifer Brett, managing director of the Upwork Research Institute in San Francisco, California, led the team behind Upwork's Future Workforce Index 2026, an annual study on how AI is changing the value of skilled work. The researchers found that freelancers doing AI-related work earn 34% more per hour than those who don't. But complexity, not just AI use, now determines who gets paid more for it.
A new role takes shape between tools and outcomes
Upwork's research team has been tracking a pattern in job posts and client requests on its platform over the past year, and gave it a working label, AI orchestrator. Brett said it isn't a title candidates would find in an actual job listing. Rather, it's a grouping her team applied to describe a skill set companies are increasingly looking for as they try to get real value from AI.
"I don't think the AI orchestrator is actually going to be like a new job title per se. It's more about the person who's figuring out how do we bring together human talent and AI to work together that's going to effectively solve bigger problems," Brett said.
She separates the role from prompt engineering, which she considers a narrower, task-specific skill.
"The AI orchestrator from our perspective is somebody who is actually connecting different workflows and parts of the business together and connecting potentially different tools, whether they are standalone AI tools or incorporating AI into existing tools, and then actually getting that workflow to work more efficiently and to work seamlessly," Brett said.
She offered a marketing example. A content team might bring in an orchestrator to map where AI and humans each enter the process, from ideation through drafting and publication. That kind of design work reflects a shift HRD has tracked, away from automation for its own sake and toward the deeper augmentation companies need to future-proof through AI.
Freelancers are moving faster than full-time employees
The Future Workforce Index found a split forming inside AI-related work itself. Straightforward execution tasks, like drafting a social post or an initial blog draft, are growing in volume on Upwork's platform, but earnings for that category fell 28% year over year. More complex work, the kind that requires embedding AI into an existing workflow, is paying off instead.
Freelancers doing it saw earnings rise 45% year over year. PwC's 2026 Global AI Jobs Barometer found a comparable two-track pattern across the broader labor market.
Brett said freelancers tend to develop this kind of skill first because of how freelance work operates.
"If you are a freelancer, you're often meeting early demand from organizations who have not seen that skill evolve within their company, and maybe they're unsure whether they will need to. So the investment to go and train and redesign roles and make a commitment to that degree can be hard for organizations to do," Brett said.
That doesn't mean full-time employees are waiting around, Brett added, even without formal backing from their employer.
"Full-time employees are using AI. We'd seen in the report it's something like more than half are using shadow AI. So even though their organizations are not even there with them, there's obviously a desire for full-time employees to learn and to use these tools," Brett said.
It's a dynamic HRD has covered in reporting on freelancing going mainstream as employers rethink hiring.
What to look for when hiring an AI orchestrator
Brett said the skill set behind an effective orchestrator combines two distinct capabilities.
"You have a hard set of AI skills, and that is that capacity to utilize the tools, to know how to manage them. But really where the AI orchestrator role steps in, it's that capacity to combine those hard AI skills then with what we would call the core human-centric skills. So that is the capacity for judgment, for effective decision-making, for tying technology to an actual business problem," Brett said.
She described orchestrators as the people who keep AI tools from multiplying without a plan.
"It's someone who has the judgment to say, this is actually the tooling that you need. Here's how you incorporate it into existing systems, and here's how it's actually going to deliver value for your business," Brett said.
Employers weighing how to source this expertise can find useful context in HRD's reporting on how the AI hiring surge is forcing employers to rethink workforce plans.
Shadow AI points to an unmet need for support
Brett's earlier mention that more than half of full-time employees are using shadow AI points to something bigger than individual curiosity. Roughly two-thirds of full-time workers surveyed for the Future Workforce Index said they wanted more support from their organization than they're currently getting.
"It does seem to be lack of support. More than two-thirds of full-time workers in the survey said they wanted more support from the organization that they don't feel they're getting today," Brett said.
She sees an opening for orchestrators to close that distance between employee experimentation and company policy, particularly as unsanctioned AI use keeps climbing, a trend HRD has tracked in reporting on shadow AI rising as employees outpace workplace controls. In practice, Brett said that work often looks like a short-term project rather than a permanent hire.
"You could see someone who has that level of expertise around some tools being able to make recommendations over what to use for your given organization and the stage that you're at, and how to best implement it, and then help to increase adoption across the organization, which could be done through training or some other means," Brett said.
"That could be an example of where an AI orchestrator is needed to actually help an organization counter that 55% shadow AI that we're seeing," Brett said.
Brett described the orchestrator as an emerging capability rather than a settled category. The practical takeaway is less about posting a job titled "AI orchestrator" and more about recognizing this skill set already, in hiring criteria, training investment, and internal promotion decisions.
