EU AI REGULATION
The AI Act, article by article
INTENTIA+ complies with the EU Artificial Intelligence Act. This page goes through the articles that matter most to an organization working with AI and shows, for each one, the part of INTENTIA+ that addresses it.
- Regulation Regulation (EU) 2024/1689
- Where Servers in Europe
- Deciding Always a person, with a second factor
- Trail Chained and signed every night
Who is who in the regulation
The AI Act splits obligations between the provider, who develops an AI system, and the deployer, who uses it in their work. When your organization uses AI to make decisions, it is a deployer.
Many obligations apply to high-risk systems, such as those involved in recruitment, creditworthiness assessment or access to essential services. If your use is not high-risk, the same pieces are what let you explain what was decided with AI.
The articles
AI literacy
That the people using AI in the organization have sufficient knowledge, given their role and the context of use.
- AICA is on every screen and explains how what you are looking at works, using the user guides.
- Turn details shows which documents the AI read and why, which rules governed it and what it cost.
Record-keeping
That high-risk systems automatically log what happens while they run, so it can be reconstructed later.
- Every turn keeps what was asked, what the AI read, which rules applied and what it proposed.
- The Audit log, the Activity feed and the history of each decision are chained line by line and signed every night.
- The record cannot be changed: not from the screens, not from the server, not by writing to the database.
Human oversight
That people can oversee the system effectively: understand what it does, avoid over-relying on its output, disregard or reverse it, and stop it.
- Decisions wait in Review until a person approves, rejects or amends them, with a second factor.
- A machine never decides. It is a rule of the permission engine and the database.
- "Pause everything" stops a workspace's connections; "Switch off and cut everything" revokes it all, with its reason.
Accuracy, robustness and cybersecurity
That high-risk systems perform reliably and withstand errors, faults and attempts at manipulation.
- Each workspace is isolated, and permission is checked on every request.
- Service and agent keys are for a single workspace and always expire.
- Facts, decisions and documents keep a fingerprint that is checked every time they are read.
Obligations of deployers of high-risk AI systems
Among other things, assigning human oversight to people with the necessary competence, training and authority (paragraph 2) and keeping the logs the system generates (paragraph 6).
- Each person has a role: who administers, who edits, who reviews and who only views.
- Each machine's autonomy is written by a person: what it does without asking, with approval or never. Each approval covers a single action.
- Settings → Export gives you your data in a file that opens in any browser, without INTENTIA+.
Transparency obligations for providers and deployers of certain AI systems
That people know when they are dealing with an AI and that AI-generated content can be recognized as such.
- Every text written by an AI says so, on screen and in whatever is downloaded, copied or exported.
- Whatever the AI looks up on the internet comes with its sources and their links.
Right to explanation of individual decision-making
That anyone affected by a decision based on a high-risk system can obtain a clear explanation of the role of the AI and the main elements of the decision.
- Each decision keeps what was asked, what was decided, the reasoning and its references.
- "See where it comes from" shows who proposed it, who decided it, under which rules and from which documents.
What your organization decides
INTENTIA+ provides the pieces. These decisions belong to your organization, and INTENTIA+ keeps each one with who made it:
- Which uses of AI it has and which of them are high-risk.
- Who oversees each use and with what training.
- Which context rules apply and who approves them.
- How long each record is kept.
Frequently asked questions
Does INTENTIA+ certify that my organization complies with the AI Act?
No. INTENTIA+ complies with the regulation as a product and gives you the record-keeping, human oversight and transparency it asks for. Your organization's compliance depends on how it uses AI, and it assesses that with its own advisers.
How do I know whether my use of AI is high-risk?
It depends on what you use it for. Annex III of the Regulation lists the high-risk uses, including recruitment, creditworthiness assessment and access to essential public benefits and services.
Where is my data?
On servers in Europe. Your text goes only to your AI provider, on your account and under your terms.
Which AI does INTENTIA+ use?
Yours. Each organization connects its own OpenAI, Anthropic or Google account, with its contract and the spending cap it sets. INTENTIA+ does not resell AI or keep a backup key.
Official text of Regulation (EU) 2024/1689 on EUR-Lex ↗
This page is for information and is not legal advice.
Keep me posted →
Security and data →
INTENTIA+ compared with an AI assistant on its own →