Enterprise telemetry volumes are rising as companies expand their use of AI, and many expect Agentic AI to accelerate that growth. An Omdia survey of more than 300 IT decision-makers in North America and Western Europe found that respondents anticipate an average 9.5-fold increase in telemetry from agentic AI workloads over the next two years. Nearly half expect growth of between sixfold and 100-fold.

The survey also found that the increase is already underway: more than half of respondents said their telemetry volume had more than tripled in the past year, with an average rise of 3.7 times. AI and machine-learning workloads accounted for about 43% of that growth.

Many organizations say they are not ready for the expected data loads. Roughly two-thirds described themselves as only somewhat prepared or less, while 22% said they had not started considering the data implications. At the same time, 35% reported widespread deployment of agentic AI.

Monitoring costs are already affecting projects. Among organizations running agentic AI initiatives, 69% said observability costs exceed their compute and infrastructure spending. Average annual observability expenditure was reported at $3.17 million, up 28% year over year, and 81% said they were looking for ways to reduce it.

The financial pressure has also led to delays or cancellations: 59% of respondents said their organization had terminated or postponed at least one agentic AI deployment because monitoring was too expensive. The article says cybersecurity monitoring, compliance and fraud detection were among the areas affected by paused or cancelled projects.

The article argues that conventional observability systems, which typically ingest large volumes of data before processing them, may struggle as telemetry scales. It points to telemetry pipelines that can route, filter, enrich and govern data before it reaches observability platforms. According to the survey, organizations with a pipeline were 50% more likely to feel prepared for data growth of sixfold or more; those with mature deployments were 80% more likely to have avoided operational cost challenges.

Interest in these tools is high: 97% of surveyed organizations had implemented or were considering a telemetry pipeline, and 70% were actively evaluating options. Respondents most often prioritized multi-destination routing, data reduction and sampling, and content-based routing.

Sources: TechRadar.