Measure
AI Share
Compare AI-assisted contribution with quality, delivery, and attribution confidence.
- Audience
- engineering leaders, managers
- Required role
- Manager or administrator with organization-wide Measure capability.
- Product access
- Measure
- Navigation path
- Measure > AI Share
Outcome
Report AI-assisted contribution with visible attribution confidence and outcome context.
Before you begin
- AI attribution configured
- Processed pull requests in the selected period
Read AI Share
AI Share is the portion of included change evidence attributed to AI assistance. Confirm the active date, team, and repository scope before quoting the result.
Check confidence
Review known, estimated, mixed, and unknown shares. A rising unknown share can make period comparisons unreliable even when the headline AI Share moves.
Compare quality
Pair adoption with quality, rework, cycle time, and delivery outcomes. The useful question is whether the system improves while AI use changes, not whether AI contribution is maximized.
Avoid misleading conclusions
Do not treat unknown as human work, estimated as certain, or line share as time saved. Changes in tooling or attribution configuration can move the metric without a behavior change.
Related guides
Configure AI attribution
Choose attribution signals and understand known, estimated, mixed, and unknown work.
AI attribution methodology
Understand known, estimated, mixed, and unknown AI authorship signals.
Measure overview
Use delivery, quality, and AI adoption signals without turning metrics into quotas.