Attribution granularity is the level of detail used to assign activity or cost to a user, project, request, container, or workload. Coarse attribution may show that a platform is expensive, but not which API call, agent, or process caused the spike. Fine-grained attribution adds more context, such as trace IDs, process metadata, or model usage records, so teams can link spend to the exact operation.
In cyber security, granularity matters because it helps detect abuse and contain blast radius. Attackers can hide noisy behavior inside shared services, abuse API keys, or trigger runaway retry loops that burn compute and increase cloud bills. Defenders use application tracing, cloud tags, and kernel-level telemetry to correlate resource use with real runtime behavior. Better attribution makes it easier to spot anomalous workloads, investigate suspicious automation, and enforce cost controls before waste turns into an operational or security problem.