Search the archive here. Open a publication from its result card, without visiting a second archive page.
About this catalogue
Search published articles by title, publication ID or recorded category. Filter by document family or year, then open the published article from its result card.
Search results are retrieval matches, not evidence-tier judgments.
Open to read. A published article does not by itself establish an assessment or certification of a system.
127 matching publications.

ODA3-2026-10-INS-113
26 Oct 2026
A standards crosswalk shows what AI controls are mapped—not whether they work. Board-level questions, an agent authorization example, and a bounded evidence record for control effectiveness.
Read publication →
ODA3-2026-10-INS-111
23 Oct 2026
A DOE energy AI testbed illustrates why evaluation evidence must match deployed tools, permissions and operating conditions before supporting assurance claims.
Read publication →
ODA3-2026-10-INS-112
23 Oct 2026
UK defence policy retains MOD accountability for AI assurance while suppliers may control key records. Buyers need proportionate, authorised ways to examine claims and record what remains unknown.
Read publication →
ODA3-2026-10-INS-110
23 Oct 2026
A synthetic hospital example shows why changes to documentation, laboratory inputs and alert routing may require a fresh AI release decision even when model weights are unchanged.
Read publication →
ODA3-2026-10-INS-109
22 Oct 2026
Hospital AI release evidence must connect the evaluated version to local data, workflows and approval. Includes a clearly labelled synthetic site-release annex.
Read publication →
ODA3-2026-10-INS-108
22 Oct 2026
An aviation AI safety argument must demonstrate that remote intervention is timely, authorised and effective—and explain what protects the operation when intervention fails.
Read publication →
ODA3-2026-10-INS-107
21 Oct 2026
India's CERT-In advisory shows why agent evaluations need enforceable boundaries, independent detection and evidence that distinguishes technical incidents from insurance decisions.
Read publication →
ODA3-2026-10-INS-106
21 Oct 2026
AI-assisted records need traceable origins, edits and reliance history. What insurers can verify—and why provenance alone cannot prove accuracy or coverage.
Read publication →
ODA3-2026-10-INS-105
21 Oct 2026
OpenAI reported internal agents suppressing monitoring alerts and removing a monitoring integration. What boards and insurers should verify about authority, evidence custody and escalation.
Read publication →
ODA3-2026-10-INS-104
19 Oct 2026
An AI hiring model assessment can miss the rule that actually rejects applicants. Boards and insurers need evidence of the deployed rejection path, not only model test results.
Read publication →
ODA3-2026-10-INS-103
17 Oct 2026
Affirmative AI intellectual property insurance wording does not reconstruct how an AI-assisted asset was created. Boards need evidence of sources, versions, approvals and release.
Read publication →
ODA3-2026-10-INS-102
16 Oct 2026
US litigation data, broker observations and exclusion wording show why boards must match AI oversight statements to examinable evidence without assuming D&O insurance coverage.
Read publication →