Dive Brief:
- Increasing agent workloads have become costly for enterprises, running up the tab on data processing and analytics, a survey of more than 500 U.S. IT and data leaders found. Three-fourths said their costs are increasing, according to the Tuesday report conducted by Ocient in partnership with market research firm Propeller Insights.
- About one-quarter of tech leaders said that their AI demand has outgrown their existing infrastructure, and and 41% of executives identified unpredictable costs as a top data challenge, up from 35% two years ago.
- Because of growing costs and demand for resources, tech chiefs must learn where and when to pull back on AI use, the report said. “Nearly all are taking action aimed at improving the bottom line: cutting what isn’t working, backing what is, and refocusing on the data underneath it all,” the report said.
Dive Insight:
Agentic systems and autonomous AI agents have increased the complexity of data processing and infrastructure for enterprises as they pursue ROI.
Tech leaders reported a higher volume of queries, more data processing needs, low latency demand and higher infrastructure costs associated with agent use, according to Ocient’s report. Global AI spend is set to nearly double in 2026, partly driven by soaring infrastructure demand, Gartner data found.
Data foundations play a large role in an enterprise’s readiness to successfully deploy AI, but only 8% of enterprises say their data is ready, according to a recent Dun & Bradstreet survey. Project failure rates have jumped in the last few years due to poor data readiness, rather than technology limitations.
Fewer than one-quarter of tech leaders in Ocient’s survey said they can support growing AI workloads without making significant changes.
Operational efficiency was where leaders reported seeing the most value, not employee productivity, the report said. A recent SAP report had similar findings — the technology doesn’t drive cost efficiency or employee productivity, as much as it’s helping employees create insights, make decisions or interact with customers.
Because of the rising costs and strain on resources, tech leaders will need to decide what’s worth funding, the report said. Only 38% of leaders said they’ve seen significant ROI in their AI investments, and nearly all are planning to improve the bottom line by cutting what isn’t working.
Nearly one-quarter of leaders said they’ve reduced or paused underperforming initiatives, and about 45% said they’ve increased investment in successful initiatives. Security risk, financial return, accuracy and cost at scale carry nearly equal weight in how leaders decide which AI projects to keep, the report said.
Tech leaders said they’re adding security or governance requirements or introducing AI usage and spending controls as they try to reel in costs and decide what to further invest in.
“Pruning isn’t pulling back — it’s making room for what’s worth growing,” the report said.
