This October, the 19th Shared Services & Outsourcing Week (SSOW) Autumn took place in Amsterdam. For three days, the city offered more than just stroopwafels, as it buzzed with bright ideas and forward-thinking discussions.
From boardroom conversations to keynote stages, the topic of Agentic AI is everywhere. It’s being branded as the next frontier in transformation, and, for GBS, the promise is incredible: autonomous agents that can learn, adapt, and manage workflows with minimal human intervention.
But amid the buzz, questions remain:
Are GBS leaders at risk of falling for another overhyped technology trend? Is the potential of Agentic AI stuck in the future? What’s the catch?
“This industry has given lip service to changing the fundamentals for years. Do you think this time around we’ll actually put in the muscle to change?”- Deborah Kops.
The G6 Digital Debate, which never shies away from the truth, covered this very topic. The panel consisted of:
- Steve Rudderham, Head of Global Business Services, Carrier
- Deborah Kops, Founder and Principal, Sourcing Change
- Imran Ahmed, Head of Europe, Business Process Services (BPS), Wipro
- Josh Peters, Partner, McKinsey & Company
- Mohib Yousufani, Partner and Senior Global Leader of Transformation & Digital, PwC
- Sesank Kandalum, Management Consultant, BCG
- Candice Sieg, Partner, Global GBS and Outsourcing Advisory Lead, KPMG
So, before diving headfirst into the Agentic AI wave, here are some candid truths to keep in mind.
7 Candid Truths About Agentic AI in GBS
1. Realistic Time Expectations Are Crucial
Agentic AI may deliver 25–30% productivity improvements in targeted use cases, but full-scale transformation will take years. Human adaptation, organizational resistance, and change fatigue are real barriers.
“Agentic AI is not a silver bullet. If your goal is to get more cost out in the next 18 months, this is not the lever for you.” – Josh Peters
2. It’s Not Just About the Tech
“If you just deploy thousands of agentic AIs into the ecosystem, that's not going to give you more than 5-10% benefit. You've got to fundamentally tackle process steps, look at your end-to-end process design across functions, then overlay digital integration.” -Sesank Kandalum
Deploying AI agents in isolation won’t cut it. Success demands a three-pronged transformation:
- End-to-end process redesign
- Integrated technology platforms
- New operating models
“It'll finally force us to break down the silos. There are functional silos, and that's going to be a fundamental shift. You've got to look at the entire end-to-end value chain, then enable by digital. - Candice Sieg
3. Agentic AI is Not RPA 2.0
Unlike RPA, which follows rules, Agentic AI can learn, adapt, and make decisions. It’s a digital colleague, not a task bot. But that also means it’s more complex to govern and integrate.
“This is the technology that is actually rewriting the rules. It doesn't need instructions anymore. It can have a brain of its own, it can learn, it can adapt, and it can make decisions for you. It will not do work; it will own workflows.” - Mohib Yousufani
4. Operating Models Will Be Upended
Expect a shift from people-dependent processes to agent-orchestrated workflows. This could reduce headcount but increase demand for new skills such as data governance, orchestration, and AI oversight.
“Agentic AI will fail if it’s just thrown at the same old issues. It has to be integrated with the right operating model.”- Mohib Yousufani
5. BPOs Must Reinvent Themselves
Some BPOs are looking to offer integrated packages, combining process expertise, platforms, and AI capabilities. But they’ll face stiff competition from consulting firms and internal enterprise teams as in-house capabilities grow.
“With the BPO partner, you're getting a package to deliver. We are bringing the process layer, orchestrators, and Hyperscalers as a package where we'll run the process, deploy the layer, and execute it as well.” – Imran Ahmed
6. Change Management is the Make-or-Break Factor
The success of Agentic AI implementation heavily depends on effective change management and bringing the organization along the transformation journey, not just the technology deployment itself.
“We’re probably going to slow down the evolution of going truly agentic, and that’s okay. That’s just the nature of human beings…Technology is not the problem. It’s ready to go. But we as humans don’t change and evolve as fast.”- Candice Sieg
7. GBS Value Metrics Must Change
Agentic AI enables GBS to move beyond traditional cost optimization metrics to deliver real-time insights, enhanced customer experience, and 100% transaction sampling, fundamentally changing how value is measured and delivered.
“Now, instead of sampling 10% of transactions, maybe we can sample 100%. That’s a dramatic unlock in quality.”- Josh Peters.
Final Thoughts
Agentic AI is not a silver bullet, but it could be a powerful catalyst. GBS leaders must approach it with clarity, discipline, and a healthy dose of skepticism.
Are you ready to lead the transformation, or will you be left navigating another tech reality crash?