Search papers, labs, and topics across Lattice.
This report establishes an evidentiary-adequacy criterion for assessing whether runtime records from agentic AI systems can support legally operative findings regarding specific events. It argues that for a record to be deemed adequate, it must not only categorize events correctly but also provide essential relational context, such as authority and temporal validity. The findings challenge the assumption that tamper-proof logs and generic frameworks alone can fulfill oversight obligations under the EU AI Act, emphasizing the need for a more nuanced understanding of evidentiary requirements.
Runtime records alone can't ensure legal accountability; they need to convey the right context to be deemed adequate for oversight.
Agentic AI systems generate runtime records, logs, traces, and audit artefacts, but the existence or integrity of such records does not by itself establish that legally operative oversight findings can be recovered from them. This technical report defines an evidentiary-adequacy criterion for a bounded class of determinations: binary findings of fact about specific events and their relations, such as whether protected data crossed a boundary, whether a human could intervene, whether an information barrier held, or whether delegated authority was valid at the moment of use. The criterion states that a runtime record can answer such a determination only if it carries both a typing that maps recorded events to the legally operative category and the relation, such as provenance, authority, derivation, or temporal validity, on which the determination's truth depends. The claim is one of necessity, not sufficiency. The report instantiates the criterion against selected EU AI Act oversight obligations and explains why tamper-proof logs, generic process frameworks, and provenance structures alone cannot establish the relevant findings. It further relates the argument to requisite variety, the Good Regulator Theorem, and the trace-versus-hyperproperty boundary of runtime verification. Companion materials and the experiment protocol are archived on Zenodo.
