Human contribution attribution
What the student originated
Questions asked, ideas introduced, claims rejected, evidence supplied, judgments made.
Live contribution telemetry
Contribution metric · Collaboration · Idea provenance
Don't ask only whether AI was used.Ask what the human contributed.
Con-Met AI replaces retrospective AI detection with an environment where the process of AI-assisted creation is observable from the beginning — and turns that process into a concise profile with clickable evidence.
◆ATTRIBUTION ENGINE ONLINE◆EVIDENCE TRAIL INDEXED◆PROVENANCE GRAPH SYNCED◆RUBRIC LAYER STANDBY◆TEACHER JUDGMENT PROTECTED◆0 GRADES INFERRED◆ATTRIBUTION ENGINE ONLINE◆EVIDENCE TRAIL INDEXED◆PROVENANCE GRAPH SYNCED◆RUBRIC LAYER STANDBY◆TEACHER JUDGMENT PROTECTED◆0 GRADES INFERRED
Start here
How it works, in six moves
Follow the path in order, or jump to the stage you need. Every page carries a next-step link, so you never have to guess where to go.
- 01 · Set upTasksTeachers assign the task brief
- 02 · Do the workWorkspaceStudent works with the AI partner
- 03 · AttributeStep 1 · AttributionObjective contribution attribution
- 04 · Define qualityRubricsTeacher builds or uploads a rubric
- 05 · EvaluateStep 2 · QualityTeacher-owned academic judgment
- 06 · RecordReportTwo-step record, side by side
AI contribution attribution
What the model supplied
Counterarguments, structure, synthesis, phrasing — recorded rather than hidden.
Emergent contribution attribution
What the dialogue produced
Ideas that existed in neither party's opening position and appeared through interaction.
Con-Met AI · Make thinking visible
Measure contribution. Trace provenance. Support judgment.
Con-Met does not determine how good a contribution was. It shows where contribution came from and how it developed.
The system measures contribution. The teacher evaluates learning.
iWhat does the Contribution Metric measure?Expand
Con-Met measures the degree to which the human, the AI, and their interaction contributed to specific functions within the creation process.
Contribution scores are not grades and do not measure quality, intelligence, effort, correctness or academic achievement. They represent contribution attribution based on observable interaction evidence.
Con-Met measures contribution. The teacher evaluates learning.
Two steps, deliberately separate
Students work on tasks a teacher has formalised. Step 1 records contribution objectively. Step 2 assesses the quality of the work against the teacher's own rubric. Neither step adjusts the other.
Step 1
Contribution metric
Objective attribution of who contributed what, from the recorded interaction. Same rules for every student.
Step 2
Quality assessment
The teacher's rubric, applied to the work itself. Separate stage, separate judgment, separate record.
Open quality assessment →The metric is not a black box
Contribution is multidimensional, so Con-Met AI never reduces a student to a single percentage. Every dimension carries a link straight to the moments in the interaction record that justify it: metric → evidence → conversation → context.
- Assignment, workspace and AI partner in one environment
- Provenance gaps flagged as questions, never as accusations
- Students see their own profile while they work
- Teacher judgment stays central — the system supplies evidence
Sample contribution profile
Understanding the task
Question prompting
Original ideas
Research / input
Critical evaluation
Idea development