Designing an Enterprise AI Control Plane: A Design-Science Reference Architecture for Policy Enforcement, Observability, and Accountability Across Heterogeneous and Agentic AI Deployments

Supplementary material for a design science research article proposing an Enterprise AI Control Plane, a reference architecture for policy enforcement, observability and accountability across heterogeneous and agentic AI deployments. The manuscript is under review at Information Systems Frontiers. The record contains two files. S1 — Per-tool capability scoring sheets (supplementary_scoring_sheets.md). Scores for eight AI governance tools (NVIDIA NeMo Guardrails, Guardrails AI, LangSmith, Langfuse, Kong AI Gateway, Portkey AI Gateway, Microsoft Azure AI Foundry, Amazon Web Services Bedrock Guardrails and AgentCore) against nine capabilities, each scored as full, partial or absent from public vendor documentation retrieved on 3 November 2026, with an evidence sentence and source URL for every cell. S2 — Meta-requirement coding sheet (supplementary_coding_sheet.md). Derivation of fourteen meta-requirements for AI governance infrastructure from the EU Artificial Intelligence Act (Regulation (EU) 2024/1689 as amended by Regulation (EU) 2026/1744), the NIST AI Risk Management Framework 1.0 with its Generative AI Profile, and ISO/IEC 42001:2023, giving for each requirement the governing clauses, a paraphrase of their demands, the supporting literature and a derivation note. Both files are plain-text Markdown. The scores reflect documentation at a single date and will age as the products change.