AI Has Rendered Companies Unable to Manage Logs, According to Dynatrace

In case it was needed to add another area where AI is causing problems, the latest study The State of Log Management 2026 by Dynatrace highlighted how AI is having a very negative effect on log management in companies. So much so that the company goes so far as to state unequivocally that "AI is breaking business log management."

Not Even Logs Are Immune to AI

One of the promises of AI was to simplify many activities, including monitoring and managing logs. In theory, AI is indeed perfect for such tasks, where one needs to monitor a constant flow of information to extract anomalies—what are referred to as outliers in English and statistical language, meaning values outside the norm.

The reality, however, appears to be different, partly because AI is not always used to analyze logs, and meanwhile, applications that make use of it produce more logs. Many more. The Dynatrace study, conducted on 450 leaders from large companies around the world, found that AI has increased log volume by 93%. The fact that companies use, on average, seven different tools to manage logs doesn’t help at all, and indeed 80% state that the attempt to transform this data into useful information is having a negative impact on both customer experience and the timelines of AI project implementation. According to the research, about a third of companies pay for observability features that are redundant or underutilized.

Companies are so overwhelmed by the amount of data that 86% is discarded to manage costs and system limitations, despite spending an average of $2.5 million per year on log management solutions.

About three-quarters of respondents say a comprehensive platform-level approach to AI-related log management is needed. 81% believe that integrating logs into observability systems and processing them must be open and automated to enable real-time analysis.

"AI is accelerating innovation in companies, but most log management systems have not been built for the scale, speed, or complexity of AI-based environments," said Mala Pillutla, Vice President of Log Management at Dynatrace. "Since AI agents operate probabilistically, treating logs, metrics, traces, and events as separate signals is no longer feasible. To make AI systems reliable and trustworthy, organizations need a unified and intelligent approach that brings together all telemetry in real time, enriching it with deep context to enable informed decisions."

The complete study is available at this address.