EU AI Act System Register — AI Inventory Tool

AI Systems Register

All data is stored locally in your browser. Nothing is sent to any server.

🤖

Free to try · EU AI Act

AI System Register

Build your EU AI Act system inventory — classify each AI system, check Article 5 prohibited uses, and generate an Annex IV technical record. 100% in-browser, nothing stored server-side.

Article 5 prohibited practices gate check
Annex III high-risk classification
Annex IV technical documentation record
Exportable register — free with a free account

🔒 All data stays in your browser. Your AI system inventory is stored locally — nothing is sent to or stored on our servers.
ℹ️ Classifications are guidance only and do not constitute a formal EU AI Act compliance assessment.

Free account required · All data stays in your browser

Step 2 of 5 — System Details

System Details & Your Role

Provide the core details about this AI system and your organisation's role under the EU AI Act.

System name *

Vendor / Developer *

Version / Release (optional)

Deployment status

Date first deployed (optional)

Intended purpose — what does this system do? *

Your organisation's role

Under the EU AI Act, obligations differ depending on your role in the AI value chain.

Provider

You developed or substantially modified this system and place it on the market or put it into service.

Deployer

You use a system provided by someone else in your own context and under your own responsibility.

Importer

You bring an AI system developed outside the EU to the EU market.

Distributor

You make a system available on the EU market without being the provider or importer.

Both provider and deployer

e.g. an internal tool you built and use within your own organisation.

Step 3 of 5 — GPAI Check

GPAI Model Check

General-Purpose AI models (GPAI) are trained on large amounts of data and capable of performing a wide range of tasks. They are governed by Articles 53–55 of the EU AI Act, separately from AI systems.

What is a GPAI model? Examples include large language models (LLMs), multimodal foundation models, and code generation models that you make available to third parties or integrate into systems placed on the market.

Is this a GPAI model (e.g. an LLM, multimodal foundation model, or code generation model) that you are making available to others?

Yes

Not applicable

What is the approximate training compute used? (FLOPs)

Systems exceeding 10²⁵ FLOPs are classified as systemic risk GPAI models under Article 51, triggering additional obligations.

Less than 10²³ FLOPs

10²³–10²⁵ FLOPs

More than 10²⁵ FLOPs

Unknown

Step 4 of 5 — Risk Classification

Risk Classification — Annex III Check

High-risk AI systems are listed in Annex III. Review the categories below and determine whether your system falls into one of them.

1 Biometric identification and categorisation ⌄

Remote biometric identification systems; systems for emotion recognition; biometric categorisation systems assigning people to categories based on sensitive attributes.

2 Critical infrastructure ⌄

AI used as safety components in management and operation of critical infrastructure including road traffic, water, gas, heating, electricity, and rail.

3 Education and vocational training ⌄

Systems used for admission or evaluation decisions in education, student assessment, proctoring during exams, and learning assessment tools that determine access to educational opportunities.

4 Employment, workers management, and access to self-employment ⌄

AI used in recruitment and CV screening, task allocation, performance monitoring, promotion decisions, and termination of work-related contractual relationships.

5 Essential private and public services ⌄

Credit scoring, insurance risk assessment, benefits eligibility, emergency services dispatch, public assistance allocation decisions.

6 Law enforcement ⌄

Polygraph/lie detectors, individual risk assessment for criminal offending, crime analytics, profiling, facial recognition in law enforcement investigations.

7 Migration, asylum, and border control ⌄

Risk assessment of migrants and asylum seekers, automated document authenticity verification, examination of applications for asylum, visa, or residence permits.

8 Administration of justice and democratic processes ⌄

AI used to assist courts and judicial authorities in research and interpretation of facts and law; AI used to influence elections or voting behaviour.

Does your system fall under any of the above Annex III categories?

Yes

Unsure

Annex III category

Does your system interact directly with natural persons (e.g. chatbots, voice assistants, generated content)?

If yes, transparency obligations under Article 50 apply (Limited Risk). If no, the system is Minimal Risk.

Yes

Step 5 of 5 — Documentation

Documentation & Human Oversight

Complete these fields to create your Annex IV-aligned technical record. This information is stored only in your browser.

Intended purpose (brief)

Pre-filled from Step 2 if entered above.

Training data description

Human oversight mechanism

Person responsible for this AI system

Review / reassessment date

Internal notes (optional)

System Record

AI System Record

Review the record below. Click "Add to Registry" to save it to your local registry.

Informational use only. This tool is provided for awareness purposes to help businesses understand their current situation regarding EU regulations. Risk classification and Article 5 checks are performed using rule-based logic built on the text of the EU AI Act — no data is sent to any server, and no AI is used. Results are indicative only and do not constitute legal, regulatory, or professional advice. Verdaio accepts no liability for decisions made based on this tool’s output. Your data is stored locally in your browser only and does not leave your device. See our Privacy Policy for details on data handling.

Your EU AI Act System Register