Dashboard
Live counts and a 60-minute launch path driven by editable sheets.
AI OPERATIONS · EXCEL + WORD
Turn AI transparency into
a repeatable launch workflow.
Inventory AI uses, run ten Article 50 QA scenarios, draft and test visible notices, retain evidence, and rerun a local READY, REVIEW, or BLOCK gate after release or change.
English · macro-free Excel + editable Word + local JSON gate · sources checked 3 August 2026
01Inventory
One row per AI system, feature, or repeatable content workflow
02Route
Separate operator role, scenario, official-checker result, and escalation
03QA
Run ten final-guideline scenarios and close owner or evidence gaps
04Launch
Tailor visible copy, test placement, accessibility, and persistence
05Evidence
Retain approvals, versions, training, incidents, and review dates
06Re-test
Run the local READY, REVIEW, or BLOCK evidence gate after release or change
WHY NOW
Build the evidence trail before launch pressure compresses the decision.
- Article 50 transparency rules apply from 2 August 2026, with a limited marking transition for certain pre-existing systems
- The Commission published final Guidelines on 20 July and updated its operational FAQ on 24 July 2026
- Direct interaction, synthetic media, and certain public-interest text need distinct routing
- Visible copy and provider-side machine-readable marking are separate workstreams
- AI literacy obligations have applied since 2 February 2025
8 WORKSHEETS + PLAYBOOK + LOCAL EVIDENCE GATE
A practical operating layer around the official sources.
Start with the official checker. Assign owners, run Article 50 QA, control copy, test implementation, retain evidence, then use the included local gate for recurring post-launch review.
AI Inventory
Fifty ready rows with ownership, role, scenario, status, and review routing.
Notice Library
Controlled baseline English copy, placement prompts, checks, and sources.
Article 50 QA
Ten guideline-aligned tests for interaction, marking, deepfakes, public-interest text, review, exceptions, and transitions.
Evidence Log
Versioned proof of notices, tests, approvals, training, and change control.
Literacy Tracker
Role-relevant AI literacy completion, refresh dates, and evidence links.
Incident Log
Disclosure, marking, accessibility, misleading-output, and review failures.
Sources & Scope
Official URLs, freshness date, status notes, and explicit boundaries.
OFFICIAL-SOURCE BOUNDARY
The kit supports operations. It does not certify compliance.
Rules, guidance, standards, and implementation status can change. Re-check live official materials before launch and escalate uncertain, sensitive, or high-impact uses.
Final Commission Article 50 Guidelines ↗Scope routingOfficial checker first
Operations kitInventory → evidence
Technical markingSeparate assessment
Not legal advice, an official EU instrument, or a conformity assessment.FAQ
Know the boundary before buying.
Does this decide whether the EU AI Act applies?
No. Use the official Compliance Checker and current guidance, and obtain qualified advice for uncertain or high-impact uses.
Does the visible notice satisfy machine-readable marking duties?
No. Provider-side technical marking is a separate engineering and conformance workstream.
Is the workbook a legal template?
No. It is an operations and evidence companion refreshed against the Commission’s final Article 50 Guidelines, FAQ, and Quick Facts on 3 August 2026.
What file formats are included?
A macro-free .xlsx workbook, an eight-page .docx playbook, notice-pattern text, and a dependency-free local JSON evidence gate with two fictional records and twelve tests.
VERSION 1.2 · SAME LAUNCH PRICE
Give every AI launch one controlled operating path.
Eight-sheet Excel workbook, ten Article 50 QA scenarios, eight-page Word playbook, and a dependency-free post-launch JSON evidence gate with two fictional records and twelve tests.