The software for AI compliance and AI governance.
Clear. Structured.
Responsible.
UNAIT is the browser-based platform organisations use to record their AI applications, classify risks under the EU AI Act and document every obligation with verifiable evidence. Hosted in Germany, productive in days rather than months.
*Developing AI guidelines and building AI literacy
Common questions before you start
UNAIT at a glance
How to meet the requirements of the EU AI Act with UNAIT in a structured way, shown through the work of an AI officer.
Common situations
and what follows from them
Three scenarios in which AI governance quickly becomes relevant.
When a chatbot is more than a helper tool
Chatbots are introduced quickly, often before anyone has established which data is processed, who has access, and whether users can tell that they are interacting with AI. The AI Act sets clear rules, and compliance with them should be documented.
More on transparency obligations →From deployer to provider
An institution starts out using an existing AI tool internally only. If the system is substantially modified, placed on the market under its own name or put to a new risk-relevant purpose, the role of deployer can turn into that of provider. Considerably more extensive legal obligations then apply.
Understanding roles under the EU AI Act →AI in recruitment or examination processes
As soon as AI prepares evaluations, recommendations or decisions, for example in HR, considerably higher requirements can arise. Human oversight matters and must be documented without exception.
Understanding high-risk AI →Every function. One platform.
From the first record to audit-ready evidence, UNAIT supports the whole AI governance process in your organisation.
A central AI inventory
Every AI application in your organisation at a glance: by risk class, status and responsibility. The colour-coded overview shows immediately where action is needed.
A structured assessment
The guided assessment takes you through the risk evaluation step by step. Only the questions relevant to your case appear. Straightforward for your team to use.
Automatic risk classification
UNAIT classifies every AI application automatically under the EU AI Act, from minimal risk to prohibited, and sets out the resulting obligations by article and annex.
Audit reports at the touch of a button
Every assessment is documented so that it can be traced without gaps. One click produces a complete PDF audit report: for audits, requests from authorities and internal evidence.
Governance Quick-Check
In a few minutes the Quick-Check shows you where your AI governance stands, with a traffic-light system and prioritised recommendations. Its seven areas follow the structure of ISO/IEC 42001, which makes the check a starting point for certification as well.
Article 4 training obligation still open? See our partner training →AI assistant in the assessment
An intelligent AI assistant is on hand while you complete the assessment: it explains questions, offers orientation on classification and gives targeted pointers. It does not constitute legal advice. Responses are generated by GWDG AI models running on servers in Germany. Nothing you enter is stored permanently.
Every compliance document kept in order
UNAIT does not stop at assessing the risk of an AI application. It then guides you step by step through the matching compliance documents, drawn from a library of 41 templates covering every role and risk class under the EU AI Act. Each relevant document can be completed in a structured way in a few minutes, giving you audit-ready documentation of your AI applications.
No AI system decides your compliance
Sooner or later someone asks. A supervisory authority, a committee, internal audit or a customer: why was this AI system classified this way and not differently? What counts then is what you can put on the table.
Some tools let AI agents do this work. The result is a generated assessment that cannot be traced step by step. And a system that can answer the same input differently is itself hard to audit.
UNAIT takes the other route. Classification follows a fixed decision logic. It was written by people on the basis of the core legal texts and reviewed by legal professionals.
Why this matters for you
Public sector and academia. Article 14 of the EU AI Act requires human oversight of AI systems. Towards supervisory bodies and committees you present your own answers, not the decision of a model.
Companies. Customers, partners and tenders increasingly ask for evidence. If you are aiming for certification, ISO/IEC 42001 is the international standard for it, you need exactly this traceability. What the standard requires
Choose the model
that fits your organisation
From a structured start with the UNAIT Risk Tool, through complete AI governance with UNAIT Compliance, to certified training via our partners.
UNAIT Risk Tool
Ideal for organisations that want a first overview and wish to record and document individual AI systems in a structured way.
UNAIT Compliance
Everything in the Risk Tool plus complete AI governance: a central inventory, approval processes and audit-ready documentation.
KI-Wisskom training
AI literacy training via our partner KI-Wisskom, tailored to managers, staff and administration.
A governance structure in four steps
No complex IT project. UNAIT is web-based and productive within days rather than months.
Demo & consultation
In a 30-minute online meeting we show you UNAIT and discuss what you specifically need.
30 days free
Start with a free trial account, no credit card and no obligation.
Onboarding
We support onboarding and roll-out in your organisation personally.
Set up and structured
AI applications recorded, assessed, documented. The basis for your AI governance.
Ready for structured AI governance?
Start with a free demo and build a lasting governance structure for your AI systems.