Almost every employer says AI capability matters, but few have the mandate, the measures or the confidence to build it.
Eighty-three per cent of employers say AI capabilities are an active workforce priority but only 29% say they have built mature ones.
Those figures come from 2U's State of Enterprise Learning 2026, published on 8 October. Research firm NewtonX, working with edX Enterprise, surveyed 507 chief HR officers, chief learning officers and L&D leaders at organisations with 500 or more employees. The distance between those two numbers is the story of AI upskilling in 2026: there's plenty of priority, but not much maturity.
The cause is not a shortage of ambition but a shortage of mandate. Nearly four in five employers (78%) lack a strong mandate to lead AI upskilling, which the report defines as executive support, clarity on which capabilities to develop, and the budget and capacity to do the work. Just 26% of respondents are highly confident their L&D function can deliver the capabilities their workforce needs (2U).
Many organisations, in other words, have an ambition without a mandate. AI capability sits on the strategy slide, but nobody has been given the direction, the money or the time to build it.
In a recent conversation with an AI transformation leader, we discussed a name for what builds up when that happens: capability debt, the widening gap between an organisation's ambitions and its people's capability. We explored this idea in Completion isn't capability, where our chair, Rob Barnett, described AI capability debt as the people version of tech debt, and one that is building up faster than tech debt ever did.
The hiring market suggests the gap could widen before it narrows. Workday's Global Workforce Report, published on 5 October, found that mentions of management and leadership skills in job requisitions fell 7% between September 2025 and July 2026, and mentions of training skills fell 13%. Employers are asking for less of the very skills that would help their people through the shift.
What the leaders do differently
2U's research identifies a small group of "AI transformation leaders": the 15% of companies that combine mature AI capability with high confidence in their ability to deliver the capabilities their workforce needs. Compared with everyone else, they are 3.9 times as likely to have a strong mandate for AI upskilling. They track 1.6 times as many measures of learning success, including outcomes beyond course participation, and they integrate AI capability building with leadership development at 2.7 times the rate of other companies. Across the whole sample, 85% do not develop AI and leadership capabilities as one connected effort (2U).
NewtonX also looked at which factors were most associated with leader status. Mandate strength accounted for 59% of the combined influence of the four factors it analysed, and measurement depth for 34%. As Chris Kelly, Strategic Insights and Analytics Director at NewtonX, put it, what separates the organisations pulling ahead is "a clear mandate, the capacity to act on it, and measures that show whether learning is actually moving workforce readiness" (2U).
Mandate and measurement, in that order. Neither is a content problem, and neither is solved by buying another course library.
Where to start
If your organisation is among the 83% that have made AI a priority, four steps will tell you whether you are on the way to joining the 15%.
Turn the priority into a mandate. Ask your executive sponsor to commit, in writing, to the capabilities you are building, the budget behind them and the time people will be given to learn. If you can't get all three, you have found your first problem.
Measure the work, not the course. Pick a handful of workflows in each priority role. Record how long they take, how often they need rework and what good output looks like. Then measure again after the learning. Sentiment and confidence still matter, but they belong alongside the evidence, not in place of it.
Connect AI to leadership development. Managers decide whether a new way of working survives its first busy week. If your AI programme and your leadership programme are run separately, bring them together.
Ask the key question every quarter. What are our people doing differently? If the answer is nothing, stop and change the programme before you scale it.
Capability debt is not paid off with more completions. It is paid off when people can do their real work better, and you can show it.
That is how AcademyAI's training is built. We assess your people's AI capability before anything else, deliver role-specific learning in the flow of work, evidence improvement beyond completion, and teach people how to prompt effectively and efficiently. If you would like to see where your organisation stands, find out more about our training here.
Sources
- 2U, "New 2U Research Finds 83% of Employers Prioritize AI Skills, but Only 29% Have Mature AI Capabilities" (The State of Enterprise Learning 2026), 8 Oct 2026: https://2u.com/newsroom/state-of-enterprise-learning-2026-ai-upskilling-research/
- Workday, "Workday Global Workforce Report: AI Is Rewriting Jobs More Than It's Cutting Them", 5 Oct 2026: https://investor.workday.com/news-and-events/press-releases/news-details/2026/Workday-Global-Workforce-Report-AI-Is-Rewriting-Jobs-More-Than-Its-Cutting-Them/default.aspx
- AcademyAI, "Completion isn't capability": https://academyai.co/blog/completion-isnt-capability
