Workday: Demand for Basic AI Skills Falls 25%

The January peak, the fall of a quarter

A Workday report published on Monday 5 October 2026 measures demand for AI skills across 550 employers. All of them use the same recruiting software, so the data set stays homogeneous.

The curve traces a clean parabola. Demand for basic AI skills climbs through late 2025 and peaks in January 2026. In the months that follow it loses a quarter of its weight, as Channel Dive^[1] reports.

Over the same span a second curve rises: technical capabilities, automation and engineering, gain 51% between September 2025 and July 2026.

Two measures on the same employer population, pointing in opposite directions. The first says knowing how to use an AI tool has become an entry requirement. The second shows where the premium has moved: towards the people who build and maintain automation.

A skill loses its price when supply catches up

A skill is worth what it is rare. Prompting, reviewing an output, daily use of an assistant: all of it entered the toolkit of millions of people in eighteen months.

The market recorded that crossover in the language that belongs to it, demand. January 2026 marks the point where employers stop asking in writing for what they now take as given.

One distinction is worth holding steady. Adoption measures how many people use a tool. Maturity measures how many can redesign work around that tool, and the 51% growth on the technical side speaks to the second.

This is where the governance problem opens for anyone running an organisation. The next level exists, it is measured, it has a market price. The road to it runs through the organisation itself.

The internal ladder has stopped

The same report carries the figure that turns the parabola into a question about people. More than half of employers report a decline in moves into new internal roles, with promotions flat.

A market asking for more advanced technical skills meets organisations that have frozen the mechanism by which a person reaches those skills.

An internal role change is the cheapest form of training in existence. It teaches on the job, it keeps institutional memory, and it gives the person visible proof that training pays.

With that door closed, the invitation to be more productive lands as a one-way request. Phil Willburn, vice president people systems, intelligence and support at Workday, describes the effect in direct terms: organisations that skip the preparatory work watch their most capable people disengage in silence.

Four in five people already know what it takes

The second leg of the report is a survey of 6,000 full-time employees, run by Workday for this publication.

Almost four in five say they know which skills they need to succeed in their role. Two thirds say their employer has helped them build those skills.

The distance between those two answers is the space where the problem lives: awareness is widespread, support reaches a smaller share.

The scope of the figure is worth fixing: full-time employees, self-reported answers, a single software vendor as the collection channel. The sample describes a population of settled workers, so the reading stays cautious about other contexts.

The gap between the four fifths who know and the two thirds who get support is a figure to take to the board. It says the demand for training already exists inside the house, ready and aware.

Productivity expected, heads counted

The report also measures what decision-makers expect. Two leaders in five expect AI to raise the productivity of the people already on the payroll. 28% expect headcount to shrink.

The two expectations coexist in the same sample and describe two opposite strategies built on the same tool.

Leaders betting on the productivity of the people already there buy time and take on an obligation: a clear map of the tasks AI automates and the tasks AI assists. Willburn places that map ahead of any demand for results, alongside recognition tools for middle managers.

Workday itself is going through the transition it describes. The company cut part of its technical team inside an AI-related reorganisation, as TheStreet^[2] reports, and its fiscal quarter numbers remain public on Fortune^[3].

The point of interest for readers sits in the consistency between the measurement and the practice.

The counterargument holds in part

A fall in demand allows two readings. The first: those skills are worth less. The second: those skills have become implicit, and so have vanished from job ads.

The second reading is plausible and it changes the tone of the conclusion. An implicit requirement remains a requirement, with its wage premium absorbed back into base pay.

The limit of the source remains, and it is worth stating. The measurement comes from inside a single recruiting system, across 550 employers, and it describes demand as expressed in job ads.

Both readings lead to the same behaviour for anyone leading people. Implicit literacy and devalued literacy call for the same answer: take people to the level where the market still pays a premium, the technical one.

The growth in technical demand, 51% in eleven months, holds under both hypotheses. The structural signal sits in that curve, with its steady direction.

What the organisations closing the gap do

Willburn points to three conditions, all of them organisational design. Say openly that roles are changing. Connect that change to a strategy people recognise themselves in, then give each person real agency over their own reskilling path.

The three conditions share one trait: they concern the leaders, ahead of the people. AI adoption remains a designed condition, far more than an individual choice.

A fourth lever emerges from the evidence gathered on this desk: converting internal talent beats external hiring on cost and on speed. Reopening internal role moves is worth more than a course catalogue.

Middle managers are the point of application. Supporting a person through a role change takes calendar time and recognition tools, two resources assigned from above.

The design question for leaders

The readings of this evidence change depending on the table.

• CEO: take organisational readiness to the board as an execution variable, with internal mobility among the indicators.
• CHRO: shift the L&D budget from basic literacy towards technical skills and the design of role pathways.
• CFO: compare the cost of an internal move with the cost of an external hire on the same technical profile.
• Talent & Compensation Committee: track the internal mobility rate quarter on quarter, alongside the promotion rate.

One question belongs in the minutes. How many people have changed role inside the organisation over the past twelve months, and with what skills pathway behind them?

A number that falls, with promotions flat, describes an organisation that asks for the next level and keeps the ladder to it closed.

The technical demand curve will follow external dynamics, outside the control of those who lead. Internal mobility, by contrast, is a decision taken in-house, every quarter.