Harvard Business Review’s study found 96% of business leaders think AI is critical but only 23% have the infra to support it – and 47% say costs have already exceeded expectations
Harvard Business Review’s Analytic Services Pulse Survey found leaders in companies across every sector are pushing agentic AI from pilot to production. Yet although 96% of senior leaders say agentic AI will be critical to their organisation within two years, only 23% believe they have the strategy and infrastructure in place to support it.
The financial toll is already apparent: among organisations already deploying agents, 47% say infrastructure costs have exceeded expectations. Companies are being told they could eliminate entire tier-one security triage teams with agentic AI, only to discover they would need to quintuple their data infrastructure spend to support it. At $10 million (€8.59 million) a year already, the value proposition disappears fast.
Also, without the right telemetry foundation, “AI systems become black boxes: ungovernable, unexplainable, and ultimately unusable at scale,” according to the press release. Further, When AI agents begin reasoning and executing across systems, at machine speed, across thousands of parallel tasks, telemetry volumes can multiply by a factor of 10 or more. Dashboards built for humans typing in queries cannot keep up.
In the report, Ryan Kurt, CEO at The AI Lab stated, “Without the right infrastructure, you’ll hit a ceiling. There is absolutely no way to break through it unless you have the data scaffolding, the governance, and the integrated workflows that you need.”
New challenges, new ways of winning
For IT and security teams, the urgency is acute: vision is not the issue so much as trying to run a new generation of AI on legacy observability and security stacks that were never designed for it. Agentic AI not only multiplies telemetry volume, but it also changes the nature of what that data must do.
Unlike GenAI, agentic AI doesn’t just summarise and suggest, but autonomously plans, decides, and acts across live business systems. While legacy tools capture what happened, agents need to know why, in real time, at scale, across every system they touch.
Organisations at the leading edge of agent deployment are finding those legacy systems are struggling and even failing under the load.
The report finds that the organisations pulling ahead aren’t just buying more AI tools. They are re-architecting their data layer and treating telemetry as a strategic input rather than an afterthought; that is, they are fusing machine data with human context so agents can reason effectively. They are also choosing open, interoperable platforms that give them flexibility as the AI landscape continues to shift.
Here are the survey’s other key findings:
- 76% report telemetry data volumes have already increased due to agentic AI – 31% say volumes have doubled or more.
- 82% anticipate a financial cost to meet agentic AI’s data infrastructure demands.
- 46% cite unclear ROI and performance metrics as the leading consequence of unprepared infrastructure, the primary reason AI projects stall.
- Top adoption barriers are privacy risks for 57%, talent shortages according to 53%, followed by unprepared data architecture (53%) and workflows not configured for agentic AI (52%).
About the research
Harvard Business Review Analytic Services surveyed 351 senior business leaders in February 2026, all from organisations actively considering, piloting, or deploying agentic AI. Respondents span North America (45%), Europe (23%), and Asia Pacific (19%); 77% hold senior or executive management roles. The research was sponsored by Cribl.
Clint Sharp, co-founder & CEO at Cribl noted, “Data is growing at a 30% CAGR, budgets are not, and now AI agents are multiplying that problem by an order of magnitude. The infrastructure most enterprises built for the last decade simply wasn’t designed for the agentic workloads of the next one.
“This research validates what we see every day, organisations know they need to get ready, and the time to modernise that foundation is now, before the rest of the organisation catches up and demands they move at the speed of AI.”
T
