The reskilling deficit: training for today's AI, not tomorrow's jobs
The Conference Board surveyed nearly 1,300 workers and interviewed 35 enterprise leaders and found most AI training covers today's tools, not the reskilling that AI-driven job transformation will require, leaving over a quarter of workers with no AI training at all.
TL;DR
- The Conference Board surveyed nearly 1,300 workers and interviewed 35 enterprise leaders, released this week.
- 55 percent of workers already use AI regularly, but only 33 percent got employer-provided AI training in the last six months, and 28 percent got none at all.
- Fewer than half say their employer gives them sufficient time or tools to build AI skills: the training gap is resourcing, not awareness.
The Conference Board report, released this week and built on a global survey of nearly 1,300 workers plus interviews with 35 enterprise leaders, lands on a distinction worth sitting with: helping someone use today's AI tool is not the same project as preparing them for the job AI is about to reshape. Most organisations, the report finds, are running the first project. Almost none have started the second.
Usage is outrunning investment
| Conference Board global survey (n ≈ 1,300 workers, 35 leader interviews) | Share |
|---|---|
| Regularly use AI at work | 55 percent |
| Received employer-provided AI training in the past six months | 33 percent |
| Receive no AI training from their employer | 28 percent |
| Say their employer provides sufficient time to build AI skills | 48 percent |
| Say their employer provides sufficient tools, access, or resources | 48 percent |
More than half the workforce is already using AI day to day. A third have had any formal training in the last six months. More than a quarter have had none, ever. The gap between the first row and the rest of the table is the same shape this newsletter keeps finding: adoption arrives faster than the organisational investment needed to make it durable.
Literacy is not the same asset as capability
Matt Rosenbaum, the Conference Board's principal researcher for its Human Capital Center, put the finding in terms that should worry any executive reading the usage numbers as a win: "Many organizations have made progress introducing employees to AI, but AI literacy alone will not create business value." His follow-on point is the operational one, and it is about time horizon, not just intensity: the organisations that benefit will be those that help employees apply AI in their actual work, keep developing new capability as the technology moves, and adapt as business needs shift under them.
Training people to use this quarter's tool is not the same investment as training people for the job that tool is about to change.
That distinction explains why the two halves of the survey diverge. Introductory AI literacy is cheap to deliver and easy to report as a completed initiative. Reskilling for a role that AI is actively reshaping is neither: it requires naming what the future version of the job looks like, which most organisations have not done, and building a training path toward it, which is a multi-year commitment rather than a one-off workshop.
Closing the deficit
- Separate the two training budgets explicitly. Literacy sessions on how to use the current tool are not a substitute for reskilling toward the role AI is changing, and reporting them as one line item hides the shortfall.
- Fix the resourcing gap first. With only half of workers saying they get sufficient time or tools, more course content will not move the number; protected time and access will.
- Name the future-state job before training toward it. A training plan without a target role is literacy with better branding.
The 55 percent usage figure will keep climbing on its own. The 33 percent training figure will not, unless organisations treat reskilling as a distinct line item rather than a rounding error inside the AI rollout budget.
Related reading
- AI skills gap persists despite widening personal use (CIO Dive)
- The fluency gap: near-universal AI use, thin business skill
Source: Most Organizations Are Preparing Workers for Today's AI, Not Tomorrow's Jobs, The Conference Board.