liveBook · Manning

Data flow diagram

As explained in AI Governance

A data flow map is a description of the sources, destinations, and movement of data through a system.

A data flow map records where data comes from, where it goes, and how it moves. For AI agents, non-deterministic behavior, adaptation, and complex action chains make static, linear maps insufficient for privacy oversight. Monitoring actual data and decision flows helps maintain an audit trail. section 5.6.4

Agent monitoring can capture events such as:

  • Data accessed, including whether it is personal or sensitive, and the sources accessed. section 5.6.4
  • External calls, autonomous decisions affecting individuals, errors, human overrides, and policy exceptions. section 5.6.4

How AI Governance covers it

How deep it goes
Contrasts catalogued linear flows with adaptive agent action chains and specifies telemetry events, but gives no diagramming method.
Examples
Lists data accesses, external calls, autonomous decisions, accessed sources, errors, human overrides, and policy exceptions as monitored events.
What you can do after
Diagnose compliance regressions by reviewing agent traces and tracking spikes in attempts to use disallowed data sources.
How it connects
Links data-flow mapping to AI agents, privacy governance, observability, audit trails, and OpenTelemetry.

Also explained in

Compare how 2 books explain data flow diagram →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of AI Governance. The section links point to the parts of the book it draws on.