EU AI Act · Regulation (EU) 2024/1689

Use AI. Stay in control.

Know which AI your firm runs, classify it, train your people and keep a person in charge, next to your risks, suppliers and incidents.

Request a demo

The problem

Nobody has a list of the AI tools people already use, or what data goes into them.

No one has decided which AI Act class each system falls into, or written down why.

There is no proof that staff are trained, or that a person checks what the AI produces.

How ActiveERM handles it

  1. 1
    List
    Every AI system you build or use: purpose, owner, vendor, the data it sees.
  2. 2
    Classify
    Prohibited, high-risk, transparency or minimal, with the reasons kept.
  3. 3
    Assign
    The obligations that follow, each with an owner and evidence.
  4. 4
    Train
    Record who uses which AI and the training they have completed.
  5. 5
    Oversee
    A person approves what the AI proposes, and every decision is logged.
From the AI you already use to a record you can show
Two duties already apply

Since 2 February 2025 the prohibited AI practices are banned and every firm that uses AI must make sure its people have enough AI literacy. The dates for the high-risk obligations are being changed by the EU, so check the current date before you plan around it.

We run our own company this way

ERMAPP runs its sales research, search optimisation and software work with AI agents. They only draft. A person approves every email before it leaves, every change before it ships, and one switch stops them all. Our agents are registered and classified in our own ActiveERM, in this module.

What you get

AI that drafts, never sends

The assistant in ActiveERM can draft an email but cannot send it: the database itself blocks it. A person sends.

Answers only from what you may see

Ask answers from the records the signed-in user is allowed to open, nothing else.

An AI notice for every user

Users see when they work with AI and what not to enter. Your compliance team can edit the wording.

AI suppliers in your DORA register

Record AI vendors as ICT third parties, with contracts, the services they support and exit plans.

AI risks in your risk register

AI risks, owners, controls, KRIs and actions sit with the rest of your risks and reach the board the same way.

AI governance module

AI system register, AI Act classification with passkey sign-off and versions, obligations by class, AI literacy records, and an oversight log with a pause switch and its KRIs. See it on the demo.

Frameworks covered

AI Act (EU) 2024/1689GDPRDORAEU hosting

See it working

We walk you through it on an example close to your own set-up.

Request a demo