Dynatrace State of Observability 2025 - human-in-the-loop isn't slowing AI down, it's preventing six-figure cloud bills

Artificial Intelligence (AI) now touches nearly every enterprise system. According to Dynatrace's State of Observability 2025 report – a global survey of 842 senior technology leaders across IT, Security, Operations, DevOps, and Sustainability functions – every organization uses AI in some part of its operations. Yet the report shows that widespread adoption does not automatically translate into certainty about how AI behaves, or confidence that it will act safely at scale. Leaders are relying more heavily on observability to close that gap.

To understand how this shift plays out in real environments, I spoke with Josh Clay, Solutions Engineering Director at Dynatrace, whose team works with organizations grappling with the day-to-day implications of AI-driven operations. His insights align with the report's findings and with earlier perspectives from Dynatrace executives, who have argued that many organizations still treat observability as a passive monitoring layer rather than a way to connect system ...