AI Workforce Transformation: 78% Say AI Will Free Them Up. 32% Are Already Worried About Their Job.

74% of GBS leaders already have an AI vision for their team. Last month’s “20% Problem” post left off on how few of them, just 20%, can actually measure if that vision is working, and this time we’re back in the same AGOS-Roland Berger white paper, GBS in the Digital & AI Era, but a different section of it: Workforce Transformation, not Readiness, and it’s just as uncomfortable, since 78.2% of GBS professionals say AI will free them up for higher-value work, while 31.8% say the same conversations are already making them worried about their jobs (Source: AGOS-Roland Berger, GBS in the Digital & AI Era survey), two different questions in the same section, pointing at the same underlying tension, and most leadership decks only show one side of it, usually the optimistic one.

The problem is you can roll out the tool, hit your adoption targets, and still lose your workforce, not because the AI didn’t work, but because nobody dealt with the 31.8%.

The anxiety isn’t irrational

Almost 41% of respondents told us they’re still unsure what AI means for their role (AGOS-Roland Berger survey), not against it, not for it, just unsure, and that’s the largest group in the whole survey, the one most transformation plans ignore, because “unsure” doesn’t show up as resistance; it just shows up later, as disengagement, as people quietly protecting their own jobs instead of redesigning them.

If you’re only tracking licence usage and adoption rates, you won’t see any of this coming.

The work is changing, not just the tooling

At Air Liquide, Puneet Gupta and Lalitha Kumar, speaking specifically about the GBS industry in Malaysia, put it plainly: management jobs have evolved “from managing exceptions, to managing bots, to managing citizen developers.” That’s not a tooling upgrade, it’s a different job, three different jobs actually, inside the same title, over a few years.

We heard the same point, sharper, on our own stage at AGS26. Michelle Yong Wai Cheng (Friesland Campina & Dutch Lady) put it this way: “AI isn’t just changing the work. It’s redefining the workforce. Technical competence gets you the job. Employability is what keeps you relevant.”

That’s the part most AI rollouts skip: they train people to use a new tool and assume the role underneath stays the same, when it doesn’t, and if you don’t tell people that out loud, they feel the shift happening to them instead of with them.

What the organisations ahead of the curve are actually doing

Roche doesn’t treat this as a messaging problem: Rao Murugesh told us they give employees a real sandbox to test and explore AI solutions themselves, backed by Roche Tech University, aiming to reach 100,000 employees in Q1 2026. His read on where this goes: GBS roles get more important, not smaller, as they shift toward more value-adding work.

That’s the difference: Roche isn’t reassuring people that their jobs are safe, they’re showing them, concretely, what the next version of their job looks like, and giving them a safe place to build toward it.

As our own Te Sheng Chang puts it, on what it takes to build an AI Center of Excellence (CoE): “A well-executed AI CoE empowers organizations to move beyond pilots and into enterprise-wide impact, guided by business-led champions, upskilled talent, and structured governance.” Notice what comes first in that sentence: not the technology, but the champions and the talent.

Why this is a leadership problem, not an HR problem

Skip the people conversation, and three things happen, in order:

1. The 40.9% who are unsure stay unsure, and quietly disengage instead of leaning in.

   
   

2. Your best people, the ones who actually understand the process, start treating AI as a threat to manage instead of a tool to shape, because nobody told them what their job becomes next.

   
   

3. You end up measuring the rollout by licence usage, because that’s the only number you bothered to track, and it tells you nothing about whether anyone’s actually better off.

So, what’s the one thing to take from this

The technology conversation is running ahead of the people conversation in almost every organisation we surveyed, and closing that gap isn’t a communications task you hand to HR after the rollout: it’s the thing that decides whether the rollout actually works.

Before your next AI initiative goes live, ask one question: have we told people what their job becomes next, or have we just told them which tool to use?

Want help closing that gap before your next rollout? Our AI GBS Readiness Assessment maps where your people stand today.