Compliance logging through a single standard
The AIO 20002 helps satisfy the automatic logging, transparency, human oversight, and quality management obligations required by the EU AI Act — through a single standard.
AIO 20002 — an organization records every AI judgement in the shared vocabulary
The standard number is the goal statement. The digits 2·00·0·2 say whose values, in which domain, and which action this document covers.
Organization — teams, institutions, communities
Common — domain-independent
Base document
Record — log every judgement as hierarchy values
What this standard defines — the log grammar (tags, partial output, single-code emission), V/E/S hierarchy notation, and the per-response automatic emit format. What it does not define — which judgements are right, or any evaluation of what is recorded. A record is material for comparison, not a verdict (non-normative).
Step 2 of the integrity loop — Record
Records gain their baseline from the setting made in the previous step, AIO 20001 · Setting Workshop. The logs recorded here are analyzed in the next step, AIO 20003 · Benchmark.
Public repository — the spec and full prompts
The AIO 20002 spec and a reference implementation for EU AI Act compliance are maintained in a public GitHub repository. Licensed under Apache-2.0.
Log viewer — in preparation
A web log viewer that renders pasted AIO 20002 logs in a human-readable form is in preparation. Until it ships, the examples and full prompts in the GitHub repository cover the same ground.
AIO 20002 mapping by article
| Article | EU AI Act requirement | AIO 20002 response |
|---|---|---|
| Art. 12 | Automatic event logging — automatically records logs throughout the system's lifecycle | AIO 20002 logs are emitted automatically with every response, in a guaranteed standard format |
| Art. 13 | Transparency — provide information so users can understand the system's outputs | AIO 20002 codes are expanded into natural language and shown to users (Verification stage) |
| Art. 14 | Human oversight — an interface that lets people effectively oversee the system | Monitor population-scale distributions with the AIO 20003 dashboard |
| Art. 15 | Accuracy & robustness — measure and document system performance | Quantify alignment accuracy through AIO 20003's V/E/S distribution measurements |
| Art. 17 | Quality management system — document changes and decisions | Tracked via quarterly consensus minutes plus RFCs for AIO 20001 hierarchy changes |
| Art. 50 | Disclosure of AI-generated content — inform users they are interacting with AI | AIO 20002 logs are automatically attached to AI responses, serving as an identification signal |
Three benefits of unifying compliance with AIO 20002
- Eliminate duplicate infrastructure — instead of building a different logging system for each of Art. 12/13/14/15, consolidate into a single standard log line.
- Absorb jurisdictional differences — the same log can also address similar obligations in jurisdictions beyond the EU (Korea’s AI Basic Act, NIST AI RMF, ISO/IEC 42001).
- Lower audit costs — external auditors analyze standard-format AIO 20002 logs directly: statistical verification rather than bespoke, company-by-company audits.
Mapping to other regulatory frameworks
AIO 20002 mappings to major frameworks — the NIST AI Risk Management Framework, ISO/IEC 42001 (AI management systems), Korea's AI Basic Act, and Japan's AI Business Operator Guidelines — are in progress as separate RFCs. Proposals and participation are welcome via GitHub Issues.