Zapier, a leading AI orchestration platform, recently released findings from its AI Skills Crisis survey that revealed that while 94% of organizations already use AI in some capacity and 77% call workforce AI skills an urgent need, most companies haven’t built the training infrastructure to match that conviction.
“The urgency is real and the intentions are good, but the training infrastructure is still a work in progress,” said Emily Mabie, Senior AI Automation Engineer at Zapier. “You wouldn’t hand someone a commercial kitchen and expect a Michelin-star meal without training. AI is the same. Buying tools isn’t a training strategy. You have to help them develop the AI fluency that leads to impact.”
Key Findings
- Belief outpaces follow-through: 63% of executives view AI literacy as either mandatory (28%) or a valuable asset (35%) for their workforce, yet formal training reaches only 51% of IT and engineering teams, 43% of sales and marketing, 40% of HR, and 36% of legal.
- Learning and Development (L&D) is missing in action: Just 7% of organizations put HR or L&D in charge of AI training. Instead, 34% hand that responsibility to IT and engineering leadership.
- Companies pay for AI skills but can’t measure them: 48% of executives say AI-focused roles command a salary premium, but only 21% use formal assessments to evaluate AI effectiveness; other assessments include performance review (25%) and manager or peer input (21%).
- Confidence may be outpacing reality: 76% of executives say they have the talent to hit their AI goals, a striking number given the training gaps, unclear ownership, and fast-moving skill requirements the same respondents flagged.
Why This Matters
These findings build on Zapier’s previous research which found that untrained workers are six times more likely to say AI makes them less productive. Taken together, the data suggests that simply giving employees access to AI tools without structured training isn’t just leaving value on the table. It’s actively creating problems.
“Most organizations are stuck in a loop where they recognize AI skills are urgent but haven’t figured out who owns the training or how to keep it current,” said Mabie. “The fix doesn’t have to be complicated. Instead of pulling people into a classroom, bring the training to their work. Start with the workflows your teams already run, build AI into those, and let people learn by solving their actual problems. That’s more effective than any seminar.”
Leave a comment