AI Didn’t Come for Your Team. It Came for Your Job Descriptions.

What the Workday Global Workforce Report actually means for how companies hire in 2026.

Workday’s October 2026 Global Workforce Report analyzed job requisitions across 550 employers and found something counterintuitive: demand for basic AI skills — simple prompting, knowing how to use AI tools — peaked in January 2026 and then fell 25%. Meanwhile demand for hands-on AI skills — building tools, automating workflows, AI engineering — rose 51%.

The market stopped wanting people who can use AI. It started wanting people who can build with it.

Only 28% of leaders expect AI to reduce headcount. The other 72% are dealing with a different problem: the people they have don’t have the skills the business now needs, and the people who do are expensive and hard to find.

The Skill Gap Nobody Planned For

When companies adopted AI tools, the assumption was that existing staff would adapt. Run some training sessions, point people at new tools, absorb the change.

That assumption is proving wrong.

79% of employees say they know which skills they need to succeed. Only 66% say their employer helps them develop those skills. A 13-point gap between what people know they need and what companies actually provide.

Companies are also asking for less of the skills that would help workers through this shift: mentions of management and leadership skills fell 7% and training skills fell 13% in job requisitions.

Companies know the skills landscape is changing. They’re not investing in the transition.

The Problem With Hiring Your Way Out of This

AI expertise isn’t needed the way most technical skills are needed. A company implementing an AI agent for a specific workflow doesn’t need an AI specialist permanently. It needs one intensively during design and implementation, periodically for oversight and optimization, and urgently when something breaks or needs to be updated.

This is a project demand pattern, not a headcount pattern.

Median applicants per filled role rose to 69 from 58, while time-to-fill held near 60 days. More competition for the same roles, with no improvement in how fast they get filled. Hiring more isn’t solving the problem faster.

Keeping an AI specialist in-house permanently means paying full-time rates for a role with periodic demand. Most companies can’t justify it and most AI specialists won’t stay in a role where their skills are underused.

What Actually Works

The demand pattern for AI expertise fits the service model better than the hiring model. A service partner that maintains AI capability across multiple projects stays current in a way a single in-house hire can’t. The specialist who was considered expert-level 18 months ago may now be working with approaches that have been superseded. The field moves fast enough that ongoing exposure to varied implementations is more valuable than depth in one company’s specific stack.

The companies moving fastest on AI aren’t the ones who hired the most AI people. They’re the ones who figured out how to access AI expertise when they need it without trying to own it permanently.

Companies have moved from wanting people who can simply use AI to wanting people who can build with it. The demand is real. The supply through traditional hiring isn’t there. Something has to bridge that gap and for most enterprise organizations, that something is a service partner with the technical depth to do what the business needs and the flexibility to scale engagement up or down as the work requires.

Wamisoftware has been building software products since 2014 and working with AI in production projects long enough to see this gap develop in real time.