Employee AI Training Statistics 2026

Published October 27, 2026

Our analysts reviewed 16 worker, employer and training-industry surveys from 11 publishers, released between May 2024 and September 2026, to show how many employees receive AI training, what employers provide and how training compares with AI use as of Q4 2026.

In the following sections, we compare the share of US workers trained across surveys, what employers offer, the gap between AI use and training, training hours and spending, training by role, job level and country, and the link between training and AI use.

How Many US Workers Get AI Training

Five national surveys asked US workers whether they had received AI training. In a national survey that Google ran with Ipsos at the end of 2025, 40% of employed Americans said they used AI at work, but only 14% said their employer had offered them any AI training in the previous year. Where a survey published only the share without training, we converted it to the share with training.

Source: AI Services Companies, Employee AI Training Study, October 2026, using Pew Research Center, Google and Ipsos, KPMG and the University of Melbourne, Jobs for the Future and Gallup survey data. Data: us-ai-training-share.csv

• The two measures limited to the past 12 months, from Pew and from Google and Ipsos, both come in below 15%, while the two without a time limit, from KPMG and JFF, come in at 28% and 31%.
• Gallup's figure counts only employees who already use AI, which places it above the all-worker measures.
• The share of US workers with recent AI training is much smaller than the share who have had any AI instruction at all. Repeat surveys that keep the same 12-month window will show most clearly how that share changes.

Pew asked the same October 2024 sample whether they had taken any job training in the prior year, then asked those who had whether it was related to AI use. Multiplying the two published shares gives the all-worker rate in the first row. KPMG's US page also reports that 72% of US workers did not receive any AI training or education.

What Employers Offer

Surveys of workers and employers measure how much AI training companies provide and plan, including two years of New York Fed regional business surveys.

Source: AI Services Companies, Employee AI Training Study, October 2026, using Microsoft and LinkedIn, Federal Reserve Bank of New York, KPMG and the University of Melbourne, EY and World Economic Forum data. Data: employer-ai-training-provision.csv, ny-fed-ai-adoption-retraining.csv

• In the New York Fed surveys, the share of AI-using service firms that retrained staff was just over a third in both 2025 and 2026, while the manufacturer share rose from 14% to more than 20%.
• The same 2026 survey found that 4% of service firms had laid off workers in response to AI, up from 1% a year earlier, and no manufacturers reported layoffs in either year.
• AI use among the region's firms climbed sharply over the same period. A steady retraining share among AI users therefore means more firms in total are training staff for AI each year.

The New York Fed lists the training firms offered as basic AI literacy and tool instruction, automating routine tasks, prompt engineering and AI applications for specific jobs, and many firms stressed training on responsible use. Among adopting firms, the median share of workers using AI was 17% for service firms and 7% for manufacturers.

The Gap Between AI Use and Training

In every survey that measured both, more workers used AI or wanted AI training than said they had received it.

Source: AI Services Companies, Employee AI Training Study, October 2026, using Google and Ipsos, BCG, EY, Microsoft and LinkedIn and World Economic Forum data. Data: ai-training-demand-gap.csv

• The Google and Ipsos pair is the only one limited to US workers, and its 51-point gap measures unmet interest in training among all employed adults, users and non-users alike.
• In the same survey, the qualities employees most often rated as important in AI training were a trusted provider (77%), hands-on learning (75%), a structured learning path (72%) and experience with AI applications related to their current job (71%).
• In every pair, the share of workers who use or want AI is larger than the share with training. Many employees are therefore using AI tools with little or no formal instruction.

The Microsoft and LinkedIn row multiplies the 75% of knowledge workers who use AI by the 39% of those users who received company training. WEF's training figure adds the 29 of every 100 workers that employers expect to upskill in their current roles and the 19 expected to be upskilled and redeployed.

Training Hours and Spending

ATD reports per-employee learning hours and spending for all training, while EY and BCG report the AI-specific hour thresholds tied to higher gains and regular use.

Source: AI Services Companies, Employee AI Training Study, October 2026, using ATD 2025 State of the Industry (539 organizations), EY and BCG data. Data: training-hours-spending.csv

• ATD's cost per learning hour used rose by $42 from 2023 to 2024 (our calculation), which ATD describes as an increase of about 34%.
• None of the surveys we reviewed publishes AI-specific training spend per employee. ATD's figures cover all formal learning.
• EY's highest-gain group trained for more than 81 hours a year, at least 5.9 times the 13.7 hours ATD records for all training per employee (our calculation). Reaching that level for most staff would take a large expansion of training time.

ATD reports that 2024 direct expenditure per employee was $29 lower than in 2023, which gives the 2023 value in the second row.

Training by Role, Job Level and Country

White-collar workers, leaders and managers report more AI training or use than blue-collar workers and frontline employees, and emerging economies report more AI training than advanced economies.

Source: AI Services Companies, Employee AI Training Study, October 2026, using Jobs for the Future and BCG data. Data: ai-training-by-role-seniority.csv

• JFF's 14-point gap between white-collar and blue-collar workers in employer AI training runs alongside a 23-point gap in AI use at work (our calculation).
• Only a quarter of BCG's frontline employees say their leaders give them enough support on how and when to use AI.
• In JFF's data, employer AI training and AI use at work are both lowest for blue-collar workers. The employees with the least AI instruction are also the ones using AI least often.

Source: AI Services Companies, Employee AI Training Study, October 2026, using KPMG and University of Melbourne, Trust, attitudes and use of artificial intelligence: A global study 2025 (48,000 respondents in 47 countries). Data: ai-training-by-country-group.csv

• The share reporting AI training was 18 points higher in emerging economies than in advanced economies (our calculation).
• The US rate of 28% is 4 points below the advanced-economy average and 22 points below emerging economies (our calculation).
• A smaller share of US adults report any AI training or education than the advanced-economy average, so US employers adding AI tools are more likely to be teaching adults with no prior AI instruction.

KPMG's training measure counts formal or informal training or education in AI among all respondents, including people outside work, so it differs from the employer-training measures in Table 1.

Training and AI Use

Employees who received more AI training use AI more and report larger time savings than less-trained employees in each survey that compared them.

Source: AI Services Companies, Employee AI Training Study, October 2026, using BCG, Google and Ipsos, and EY data. Data: ai-training-and-use.csv

• BCG's regular-use rate is 12 points higher for employees with more than 5 hours of training (our calculation).
• In the Google and Ipsos survey, AI Fluent employees save a median of 5 more hours a week than AI Explorers (our calculation).
• Each survey links more training with more use or larger reported savings, but none isolates training as the cause, since employees who seek training may already be heavier users.

Google and Ipsos group AI users into AI Fluent employees, who use AI often and across many tasks, and AI Explorers, who use it on an occasional basis. EY also found that its most-trained employees were 55% more likely to leave their organization.

What Employee AI Training Data Means Going Forward

For many employees, learning to use AI at work happens on their own time, through trial and error with whatever tools their employer has switched on.

The picture will grow clearer as repeat surveys arrive. The New York Fed asks firms in its region about AI each summer, Gallup runs quarterly workforce surveys, and BCG and Microsoft with LinkedIn publish yearly workforce studies.