AI training statistics 2026

A citable index of the AI training statistics worth quoting in 2026 — adoption, the skills gap, what L&D teams actually do with AI, what the evidence says works, and how workers feel about it. This page combines our research report, The State of AI Training 2026, which reviewed ten major studies from Gallup, Microsoft, the World Economic Forum, Harvard, and across the L&D industry, with surveys published through September 2026.

The five figures that change a conversation

• Adoption is settled; readiness is not. Cite 52% use against 19% Frontier readiness whenever someone frames AI enablement as a tooling decision.
• The constraint is organizational. At 67% versus 32%, culture and enablement outweigh individual aptitude in explaining AI impact.
• L&D adopted AI backwards. 87% use it, overwhelmingly to produce content faster, while the strongest evidence sits with AI that teaches.
• Breadth of use is the cleanest ROI lever in the data — 45% of workers report gains at one or two task types, 90% at seven or more.
• Completion is not capability. With completion near 30% and 79% of companies claiming a skills-based approach, proficiency measurement is the reporting gap.

New in late 2026: what the latest surveys say

Five surveys published between April and September 2026 update the picture. The headline is consistent across all of them: use is climbing fast, formal training is not keeping pace, and the gap is widening.

Use is up, training access is down (PwC Global Workforce Hopes and Fears, 49,364 workers, 48 countries, May–June 2026)

• 64% — of workers used AI at work in the past 12 months, up 10 points year on year.
• 22% — use generative AI every day, up from 14% the year before.
• 51% — say they can access the learning and development they need, down from 59% a year earlier. Use rose ten points while training access fell eight.
• 56% — of the global workforce sits in what PwC calls the "engine room": neither scarce skills nor much AI experience. Only 40% of this group say they can get the training they need.
• 68% — of daily AI users feel confident about their job security, against 57% of infrequent users.
• 21 points — daily AI users are that much more likely to feel confident learning new skills than infrequent users.

Source: PwC, Global Workforce Hopes and Fears Survey 2026, reported by TNW, September 2026. https://thenextweb.com/news/pwc-survey-workers-ai-engine-room-training

Half the workforce is teaching itself (Cegos L&D Barometer, 5,524 employees and 498 HR leaders, 11 countries, December 2025–January 2026)

• 64% — of employees use generative AI for work; 79% use it personally.
• 49% — taught themselves AI in the past year, through experimentation or peers. Only 16% received formal company training.
• 32% — want AI training and have not received any.
• 28% — of organisations have formal, shared AI usage guidelines.
• 68% — of HR directors rank AI and automation as the transformation that will most affect skills in the next two years.
• 65% — of HR directors now prioritise upskilling over recruitment.
• 57% — use AI to personalise learning paths, up 20 points in three years.
• 41% — of employees say training arrives too late to be useful.

Source: Cegos, International Barometer on Learning & Development 2026, April 2026. https://static.cegos.com/wp-content/uploads/2026/05/27162510/PR_Barometer_LD_2026.pdf

Managers decide whether AI sticks (Gallup AI in the Workplace indicator, US employees, updated May 2026)

• 52% — of US employees use AI at work at least a few times a year; 30% a few times a week or more; 15% daily.
• 25% — say their organisation has communicated a clear plan for integrating AI.
• 36% — strongly agree their manager actively supports using AI on the team.
• 1.7x — employees whose manager supports AI are that much more likely to use it frequently, and 7.4x more likely to say AI helps them excel.
• 42% — daily AI use in technology, against 27% in finance, 22% in professional services and 9–15% everywhere else.

Source: Gallup, AI in the Workplace indicator, May 2026. https://www.gallup.com/699797/indicator-artificial-intelligence.aspx

Workers plan to learn, with or without you (Workera, 1,000 US white-collar workers, December 2025)

• 76% — plan to learn new AI skills in 2026: 40% to apply in their current role, 36% for future opportunities.
• 53% — are looking for a new role in 2026, split between external (28%) and internal (25%) moves.
• 73% — of workers who feel their employer misunderstands their skills plan to look outside, against 14% of those who feel understood.
• 57% — of workers say verified skills data should count in hiring decisions.

Source: Workera, 2026 Skills Report, January 2026. https://workera.ai/blog/76-of-americans-plan-on-learning-new-ai-skills-in-2026-workera-report-finds

What L&D is prioritising (HKIHRM Training and Development Needs Survey, 127 employers, ~80,000 employees, February–April 2026)

• 49% — of employers rank mastering generative AI as a top 2026 L&D priority, up from 23% in 2025.
• 82% — permit generative AI at work, up from 73% in 2024; 4% prohibit it.
• 40% — use AI-assisted learning tools, up from 16% in 2024.
• 19.4 hours — average training per employee in 2025, up 6.8% on 2024; training budgets average 2.5% of base salary.

Source: HKIHRM 2025/2026 Training and Development Needs Survey, reported by Human Resources Online, July 2026. https://www.humanresourcesonline.net/ai-tops-corporate-training-priorities-but-soft-skills-remain-essential-survey

AI adoption at work

Adoption is no longer the interesting question. Half the US workforce is already using AI on the job, organizational adoption is climbing faster than at any point since tracking began, and the agent layer inside everyday productivity software is growing at a rate no training function has matched. These are the numbers to reach for when someone still argues that AI at work is speculative.

52% — of US employees use AI in their role. Just over half the workforce now uses AI at work in some form — the baseline number for any AI enablement business case. Source: Gallup workplace AI tracking, 2026.

30% — use AI a few times a week or more. Regular use is a much smaller group than any-use, which is why headline adoption figures overstate real fluency. Source: Gallup workplace AI tracking, 2026.

15% — use AI daily. Daily use — the level at which AI reshapes how work gets done — is still a minority practice. Source: Gallup workplace AI tracking, 2026.

47% — organizational adoption, up six points in a quarter. The sharpest single-quarter rise Gallup has recorded since it began tracking workplace AI in 2025. Source: Gallup workplace AI tracking, 2026.

15x — year-over-year growth in active AI agents in Microsoft 365. 18x in large enterprises. The tooling is arriving inside software people already have open, whether or not training accompanies it. Source: Microsoft Work Trend Index 2026.

58% — of AI users say they produce work they couldn't have a year ago. Two-thirds also say AI lets them spend more time on high-value work — the upside case, from the people doing the work. Source: Microsoft Work Trend Index 2026.

20% — of employees don't know whether their employer has adopted AI. A communication gap that precedes the skills gap: you cannot enable people who haven't been told there is anything to learn. Source: Gallup workplace AI tracking, 2026.

The AI skills and reskilling gap

The gap between AI capability and AI readiness is the central finding of the 2026 data. Skills have not kept pace with software, and the strongest evidence says the constraint is organizational — culture, manager support, talent practices — rather than individual aptitude. That reframes AI skilling from a talent problem into a training and enablement problem.

19% — of workers reach "Frontier" AI readiness. High personal AI skill inside an organization actually set up to use it. Another 10% are skilled people stuck in unprepared organizations. Source: Microsoft Work Trend Index 2026.

67% vs 32% — organizational factors outweigh individual ones in AI impact. When Microsoft decomposed what drives AI outcomes, culture, manager support and talent practices explained 67% of the result; individual factors explained 32%. Source: Microsoft Work Trend Index 2026.

40% — of job-required skills will change by 2030. The macro backdrop: 170 million new roles created and 92 million displaced over the same period. Source: World Economic Forum, Future of Jobs Report 2025.

59 in 100 — workers need reskilling or upskilling by 2030. Of those, 11 are unlikely to receive it — more than 120 million workers at risk of redundancy on current trends. Source: World Economic Forum, Future of Jobs Report 2025.

63% — of employers name the skills gap as their biggest barrier to transformation. Ahead of budget, technology and regulation — the constraint employers report first. Source: World Economic Forum, Future of Jobs Report 2025.

77% — of employers plan to upskill their workforce for AI. Intent is close to universal. The readiness figures above show how far intent currently sits from infrastructure. Source: World Economic Forum, Future of Jobs Report 2025.

What L&D teams actually do with AI

L&D has adopted AI almost universally — and adopted it as a content factory. The top use cases all make production cheaper and faster. None of them changes what happens when a learner sits down to learn. This category is the one most likely to change a roadmap conversation.

87% — of L&D teams already use AI. Only 2% report no plans to use it at all. Adoption inside the training function is effectively complete. Source: Synthesia, AI in L&D Report 2026 (n=421).

84% — cite speed as their main incentive for using AI. The dominant motive is faster production, not better teaching — and speed saturates once everyone has it. Source: Synthesia, AI in L&D Report 2026 (n=421).

74% — use ChatGPT as their AI tool of choice. The most common AI tool in L&D is not a learning platform. That matters given the evidence on purpose-built versus general-purpose tutoring below. Source: Synthesia, AI in L&D Report 2026 (n=421).

63% / 60% / 52% / 38% — voice generation, content and quiz drafting, video creation, translation. The four leading use cases, in order. Every one of them speeds up content production; none of them changes how learning happens. Source: Synthesia, AI in L&D Report 2026 (n=421).

72% — say personalized learning is what they most want from AI. Followed by wider reach (65%), better engagement (56%) and clearer business impact (55%) — a description of tutoring, not content generation. Source: Synthesia, AI in L&D Report 2026 (n=421).

47% — of leaders say their AI training has job automation in mind. Useful context for why employee trust in AI training programs is not automatic. Source: TalentLMS, 2026 L&D Report.

Training effectiveness, measurement and ROI

The best-evidenced use of AI in learning is tutoring, and it is the use case corporate L&D has barely touched. Alongside it sits a measurement problem: belief in training is near-universal, but the evidence trail in most organizations stops at "completed." These are the figures that support moving from completion reporting to proficiency reporting.

2x — learning gains from a purpose-built AI tutor, in less time. In a randomized crossover experiment, physics students learning with an engineered AI tutor learned more than twice as much as students in an actively-taught classroom, and reported higher engagement. The tutor was built with pedagogical guardrails and supplied with correct solutions — raw chat is not the same intervention. Source: Kestin et al., Harvard AI tutoring study (n≈200).

45% → 90% — productivity gains rise with breadth of AI use. Among workers applying AI to one or two task types, 45% report productivity gains; at three to four it is 66%, at five to six 78%, and at seven or more 90%. Same tools, same companies — the difference is breadth of enablement. Source: Gallup workplace AI tracking, 2026.

~30% — typical completion rate for online courses. The industry's default success metric, set against 79% of companies claiming a skills-based approach. Source: Industry benchmark cited in The State of AI Training 2026.

95% — of HR managers agree better training improves retention. Belief in training is not the bottleneck. Proof of what training produced is. Source: TalentLMS, 2026 L&D Report.

79% — of companies say they are moving to a skills-based approach. Skills-based talk has outrun skills-based measurement — few of those companies can put a number on capability. Source: TalentLMS, 2026 L&D Report.

49% — of L&D professionals report executive worry about skills to execute strategy. Executives are not asking whether people finished the course. They are asking whether people can do the work. Source: LinkedIn Workplace Learning research.

Workforce sentiment and career development

The people being trained have a view, and in 2026 it is more ambivalent than the adoption curve suggests. Workers report real costs alongside the gains, and the connective tissue between manager, career plan and training is thinner than it was a year ago.

36% — of employees say generative AI is weakening their problem-solving. A reported cost of unstructured AI use, and an argument for deliberate unassisted practice inside curricula. Source: TalentLMS, 2026 L&D Report.

43% — of "Frontier" professionals deliberately work without AI to stay sharp. The most AI-fluent people in the workforce are practicing unassisted. Most training programs never ask anyone to. Source: Microsoft Work Trend Index 2026.

35% — of employees would leave a company that denied them training. Training access is a retention lever, not only a capability lever. Source: TalentLMS, 2026 L&D Report.

36% — of organizations qualify as career development champions. The minority of employers where development is a real system rather than an intention. Source: LinkedIn Workplace Learning research.

15% — of employees had a manager help build a career plan in the past six months. Down from 20% a year earlier — the manager layer is getting thinner precisely as the skills demand rises. Source: LinkedIn Workplace Learning research.

How to cite this page

These figures are free to quote and republish with attribution. Cite the original study where we name it, and link back to this page or the underlying report. Suggested citation: OneRange, "AI Training Statistics 2026," onerange.ai/resources/ai-training-statistics.

Last updated October 5, 2026.