88% of organizations now use AI in at least one function. Only 13% of workers have received any AI training whatsoever. That gap explains a lot about why most companies are seeing almost no return on their AI spending.
The numbers come from a compilation of recent studies, including McKinsey's latest survey data and a March 2026 Anthropic analysis of OECD countries. The picture they paint is consistent: companies bought the tools, skipped the instruction manual, and are now wondering why the results are underwhelming.
Only 6% of organizations qualify as "AI high performers" with meaningful business impact. Just 39% report any effect on EBIT (earnings before interest and taxes - basically, operating profit). For most of that 39%, the impact is less than 5%. Two-thirds of companies remain stuck in pilot or experimentation phases, running small tests that never graduate to actual workflows.
The Department-by-Department Reality
The adoption numbers look impressive until you examine what "using AI" actually means in practice.
Sales teams have 62.8% AI exposure, and 56% use AI daily. But daily use means drafting emails and summarizing calls. Strategic work like pipeline analysis and objection handling? Untouched. About 20% of sales activities could be automated today, and almost nobody has done it.
Marketing is worse. 75% of marketing teams have adopted AI tools, yet 84% still send generic campaigns. The problem isn't the tools - it's that only 58% of marketers can even access their own service data, 56% have sales data, and 51% have commerce data. You can't personalize what you can't see.
HR and recruiting shows the starkest disconnect: 87% adoption, but only 17% describe their implementation as "highly successful." Screening accuracy and bias remain unsolved. The tools are running, but nobody trusts the output enough to act on it without manual review - which defeats the purpose.
The Size Gap and the Age Problem
Company size predicts AI adoption more than any other factor. 52% of large companies use AI compared to 17.4% of small firms. In legal, 39% of firms with 51+ lawyers use AI tools versus 20% of smaller practices. The technology is identical; the difference is having someone on staff who can figure out how to make it work.
Meanwhile, entry-level workers are catching the worst of it. Workers aged 22-25 show a 14% decline in job starts within AI-exposed occupations. Companies are not replacing these roles with AI - they're just not backfilling them, hoping the tools will cover the gap. Given that those same tools are barely being used effectively, that bet looks shaky.
The 77% Lie
77% of employers claim to have workforce AI upskilling plans. 13% have actually trained anyone. That 64-point gap between intention and action is the real story here. Companies are treating AI like they treated previous technology waves - buy it, deploy it, assume people will figure it out. The difference is that AI tools change faster than anything that came before. The gap between "installed" and "effective" widens every month that training doesn't happen.
The practical takeaway for individuals: do not wait for your company to train you. Pick one task you repeat frequently, learn to use AI for that specific task, get good at it, then expand. The 13% who got formal training have an advantage right now, but it's an advantage anyone can close on their own.