Tech runs ops. Talent runs tech.
I’ve seen few descriptions that better capture the evolving organizational dynamics around AI.
The verbiage above is central to a new report from IBM’s Institute for Business Value: “Orchestrating Agentic AI for Intelligent Business Operations.”
Like many such reports, the findings (based on research across 750 companies) paint an optimistic perspective on how AI agents will automate processes, accelerate operations, and redefine human-tech interaction in positive ways.
What’s unique about this report is the outlook by business operation or function — and how the expected benefits of AI, as well as plans to engage third parties to get AI initiatives up and running quickly, vary greatly by function. To sum up that variance, data I’ll share from IBM below shows customer service and human resources setting aggressive expectations in terms of expected AI benefits, with significantly more modest benefits expected in finance and procurement.
The procurement function is notably aggressive in its plans to engage third parties outsourcing services to effect AI-centric transformation, while finance is low on that same scale relative to other functions.
High Expectations
The report includes a number of findings that you’d fully expect to see in a broad analysis of AI plans and expected impact that support the notion of tech running ops and talent running (AI) tech:
- 87% anticipate redefining how teams work due to agentic AI by 2027
- 83% of respondents expect AI agents to make proactive recommendations based on learning in the same timeframe
- 81% of agents will continuously improve their own performance by making feedback-based adjustments
- 75% say agents will execute transactional processes and workflows autonomously for 24/7 availability, accelerating operational velocity
The one potential question I’d raise about these findings: isn’t it likely that with the pace of agent development and tech advancement that these results could be realized in less than two years? Seems entirely possible to me but the findings align directionally with lots of analyses and data points being developed across the tech industry.
Unexpected Insights
To me, this report breaks new ground in both the written descriptions and graphic depictions of AI org structures — including the interactions between agents and humans — as work processes unfold. This example, while highly detailed, is also highly insightful, so I’m presenting it here for readers to chew on:
I found it unique to have people cited for “orchestration” (most analyses and software firms are listing orchestration as a software function and deliverable they offer), outcome governance, and innovation vision. And similarly unique that agents are noted to perform “impact evaluation” and “feedback loop and decision-support analyzer” functions.
In a helpful guide to unlocking agentic AI’s potential, the report recommends centering the business operating model around outcomes, not tasks, and this includes considering “management of digital labor” as a new profession. Note that others have also recommended adding agents and the functions they perform to org charts, as well as defining optimal human-to-agent ratios, which closely aligns with this thinking.
In a set of recommendations on preparing to scale securely, the report urges companies to not leave AI governance behind, pointing out that autonomous AI scales opportunity as well as risk, so governance must scale as well. Related point: recent research from ServiceNow on AI Maturity noted that AI and data governance is holding customers back from stronger AI maturity and outcomes.
Business Function Perspective
As noted above, a few of the operational functions analyzed stand out for their strong AI impact outlook, while others stand out for a less optimistic forecast.
On the optimistic side are customer service and human resources. In customer service, 56% of executives say they expect touchless customer data management by 2026, while 71% expect touchless customer support inquiries. That latter process, in particular, should represent massive cost and productivity savings if the impact is indeed that strong.
A less rosy outlook is reported by the finance and procurement functions. In finance, less than 30% of respondents anticipate touchless processes in financial reporting and intercompany transactions by 2026. In procurement, respondents expect less than 40% touchless functionality in supplier selection and risk management, as well as sourcing.
Customer service and human resources are two functions taking an aggressive stance on engaging business process outsourcing expertise (see chart), which may well account for the optimistic outlook on AI-driven performance in these areas.
Procurement, too, is taking a highly proactive stance on engaging outsourcers, a finding that makes its modest outlook on results somewhat surprising. Finance has the least ambitious plans to bring in third parties to help that function achieve accelerated AI results — and the function indeed does not expect to rapidly move forward with AI.
This report contains a vast amount of insightful data that I encourage readers to check out; in the meantime, I am seeking out one or more of the report’s authors to add more context to these findings.
Ask Cloud Wars AI Agent about this analysis
