When AI frees capacity in a role, you have four choices for the people involved. A good plan uses all four, role by role.
The options
- Upskill in place: the role stays, tasks change (mostly augmentation). People learn to work with AI and take on higher-value work in the same role.
- Reskill and redeploy: the role shrinks (a lot of automation). People move to growing roles inside the company, with training for the new skills.
- Hire: you need skills you cannot build in time (for example AI engineering, data, AI governance).
- Reduce: where demand truly falls and redeployment is not possible — with notice, fairness and support.
What employers expect
The World Economic Forum's Future of Jobs Report 2025 (over 1,000 employers) estimates that of every 100 workers, 59 will need training by 2030: 29 upskilled in their current roles, 19 reskilled and redeployed within their organization, and 11 who need training but are unlikely to get it. 85% of employers plan to upskill their workforce, and half plan to transition staff into growing roles.
Simple rules of thumb
- High share of hours freed and enough FTE to matter → plan redeployment, starting with adjacent roles.
- Mostly augmented tasks → plan reskilling/upskilling in place.
- Mostly human tasks → monitor; protect the human work.
- Skills not available internally within the time needed → hire (or partner).
Make redeployment real
Redeployment fails when it's an intention. It works when there are named target roles, a skills gap analysis, a funded program, a hiring rule that favors internal candidates, and a manager who owns each move.
A real example
After AI agents took over a large share of customer-support conversations, Salesforce's CEO said in September 2025 that its support headcount fell from about 9,000 to about 5,000. A company spokesperson said Salesforce had redeployed hundreds of employees into areas such as professional services, sales and customer success.