Jobs are being rewritten, and the way into them is closing

Leaders plan to get more out of existing people as AI reshapes their roles. The internal transfer channel that plan depends on is stalling. The fix is structural, not another entry in the training catalog.

The decision in front of the CEO and CHRO this quarter is not how much to spend on reskilling. It is whether the internal transfer channel can carry the people that spending produces. Every reskilling plan assumes trained employees can move into the roles the training prepared them for. That assumption is now the weakest link.

Workday's October 2026 Global Workforce Report states the shape of the problem. Forty percent of business leaders expect AI to help them get more out of the employees they already have. Only 28% expect it to reduce headcount. The bet is redeployment. But the channel is narrowing: moves to new roles inside the same company fell at 57% of employers, and promotions stayed flat. As Workday's Phil Willburn put it, employees may not be changing jobs, but their jobs are changing around them.

The cost of getting this wrong is concrete. Reporting on internal mobility carries a refrain that CHROs repeat so often The Josh Bersin Company has documented it: it is easier for someone to quit and get rehired than to transfer internally. A company that pays to reskill people it cannot redeploy is subsidizing a competitor's hiring.

The evidence points away from motivation and toward the structure underneath it. Only 9% of companies have a user-friendly, skills-based internal talent marketplace, and Fuel50's 2026 survey of more than 800 HR leaders found that only 25% of organizations filled more than half of their open roles internally. Headcount is budgeted by department. Managers are measured on their own team's output. An internal transfer registers as a loss with nothing in return. As Perry Timms of PTHR told Customer Experience Magazine, talent hoarding is not bad behavior. It is the design working exactly as intended.

So the decision rule runs before the training budget, not after it. Fund reskilling only against roles that carry a named internal landing path. Before the next program launches, run one test transfer end to end: requisition, pay band, approvals. If the test transfer costs the manager more time and goodwill than an external hire, the structure is where the money should go first. Training without a transfer path is a gift to the labor market.

Caveats, stated plainly. These are survey and self-reported data, and part of the stall is cyclical. Mercer reports 2027 compensation budgets flat for a fourth year, with 57% of executives saying the economy will have at least a moderate impact on compensation planning. Nearly half of employees say progression in their organization is based on tenure instead of skills. Workday sells the HR software that would instrument all of this, and Fuel50 sells talent marketplaces, so both have a commercial interest in the problem they measure. The hoarding dynamic is structural and old, which cuts both ways: it has always been the barrier, and it is now the barrier to a plan that depends on people moving.

An illustrative check: a company spending $2,500 per employee to reskill 800 people commits $2M. That figure is illustrative. Its return rests on whether those 800 can move into the roles the training prepared them for. If the internal door stays shut, the company paid to train someone else's new hire.