Incident Audit Fault Attribution: dataset for "Separating Governance Designation from Operational Fault in Model-Generated Incident Audits"

Rows and summaries of model-generated incident audits for the main control study and the dose extension, the superseded pilot allocation results that the paper retains for transparency, both study protocols, and the collection and analysis code.

Dataset accompanying the paper Separating Governance Designation from Operational Fault in Model-Generated Incident Audits, Trustworthy AI for Good Workshop (AI4GOOD) at NeurIPS 2026 (https://openreview.net/forum?id=oZDf3mJY4F).

Paper abstract. When a language model drafts an incident audit, it may conflate the party that caused a failure with the party a governance charter makes answerable for it. We test that distinction in a construct-separated, artifact-complete, model-only audit spanning fixed incidents, several governance descriptions, and three model families. The models consistently track the stipulated source of operational fault while assigning greater governance accountability to the explicitly designated bearer. Changing that bearer moves governance ratings in the prespecified direction without collapsing them into causal contribution or operational fault. Absolute AI designations reach a ceiling, however, so a frozen shared-versus-sole extension distinguishes a dose response for two model families but not for the third. Every request, response, failure, retry, parse, and protocol deviation is retained without repair or renormalization. The audit therefore shows how to test whether model-generated reports preserve a governance distinction under a declared rubric. It does not establish moral judgment, human agreement, legal validity, or deployment benefit.

Known limitations.

• Model-generated audits only; no human judgments of fault.

License: Creative Commons Attribution 4.0 International (https://creativecommons.org/licenses/by/4.0/). Source code files are released under the MIT License.