Control Evidence Object Model, version 0.9.10

A record model for what an AI agent is permitted to do, what it is expected to do, what it was observed doing, and the evidence for each. Permissions for AI agents are well served by existing standards: scopes, rich authorisation requests, token exchange and authorisation APIs all describe what an agent may do. Observed actions and the evidence for them have no common schema. The consequence is that an organisation can hold a register, a policy, a committee and a risk assessment process and still be unable to state whether what its agents did matches what anyone intended. This model records that layer. Twelve objects across four layers, from an obligation through to the evidence for it, with three positions held on every action rather than two: what the grants permit, what the accountable person expects, and what was observed. The divergence between the second and the third is where authority expands without anyone deciding that it should. The deposit contains the specification, a field dictionary, the controlled vocabularies with definitions, a JSON Schema, and a reference validator implementing the conformance rules. Annex D states what conformance means and distinguishes a defect, meaning the records are malformed, from a finding, meaning the records are well formed and say something about the subject. Conformance is the absence of defects and not the absence of findings. Status: pre-1.0. The twelve objects, the four layers, the evidence scale, the reference states and the gap scale are stable; field-level detail may still move. The model has been instantiated against two real agent assessments. The worked examples in the specification are constructed deliberately, so that no reader receives a client estate along with the document. What the model still lacks is two assessors reaching the same determinations independently from the same artefacts. The assessment practice built on this model is not published and is not required in order to implement it.