🚀 UNAIT launches on 1 October 2026
–Days
–Hrs
–Min
–Sec
EU AI Act 2024/1689

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.

Explore solutions Download brochure
Already 11 organisations have registered for the demo
app.unait.de / ai-inventory
HS
AI Inventory
28 AI systems recorded · Last export: 12 May 2026
+ New case
Create audit report ▾
1 Prohibited 4 High-risk 8 Limited risk 15 Minimal risk
Risk ▾
Status ▾
Created by ▾
ChatGPT use in customer service
High-risk ✓ Approved
›
AI-assisted candidate screening
High-risk Draft
›
Predictive maintenance – plant 3
Limited risk ✓ Approved
›
Text correction with Microsoft 365
Minimal risk ✓ Approved
›
Citizen services chatbot
Limited risk Draft
›
90%
Less time spent on risk classification
250+
Institutions supported*
30+
Years of combined software development
100%
Data stored in Germany

*Developing AI guidelines and building AI literacy

Sound familiar?

Common questions before you start

AI in organisations
Challenge 01

Do you know which AI tools are used across your organisation?

Many institutions cannot say with certainty what staff use AI applications for, or with which data. That is exactly where the grey area begins.

To meet the obligations of the EU AI Act, organisations first need to know where AI is in use and which risks come with it. Without structured records and risk classification, obligations go unnoticed. UNAIT records your AI applications, assesses risks under the EU AI Act and produces audit-ready evidence: fully automated, securely hosted in Germany, with no IT effort.

High-risk AI
Challenge 02

AI tools fall into the high-risk category faster than expected

Can you say with certainty that none of your staff use high-risk applications? And have you introduced a compliance system that documents the use of those systems for external audits?

Where using AI quickly becomes a risk
  • Is personal or confidential data being entered?
  • Is AI used for evaluations, recommendations or decisions?
  • Is it clear whether approval or documentation is required?
  • Could you produce an AI inventory in the event of an external audit?
  • Do you carry out risk assessments of your applications, as the EU AI Act calls for?
Documentation obligations
Challenge 03

Staggered obligations from 2025 – documentation becomes a core task

Missing reviews, unclear roles, unmet documentation obligations or incorrect classifications can have legal consequences. For certain infringements the EU AI Act provides for fines of up to €35 million.

Acting early creates solid evidence and saves considerable effort if it ever matters.

Explained in three minutes

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.

From everyday practice

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 →
Transparency obligation: users must be able to tell that they are interacting with AI (Article 50 EU AI Act)
Clarify: which data is processed, and by whom?
Classify: is the chatbot a high-risk system?
Document: who granted approval, and when?
UNAIT guides you step by step through review and evidence

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 →
Role clarity: deployer or provider, and which specific obligations apply to each
Substantial modifications can shift the role in legal terms
Providers carry considerably greater responsibility (conformity assessment, CE marking)
UNAIT supports role classification in the assessment and sets out the obligations that follow
A documented initial assessment as the basis for a legal review

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 →
Employment decisions fall under Annex III of the EU AI Act (high-risk)
Human oversight must be active and documented
A risk management system, technical documentation and logging are required
Affected persons have a right to an explanation of the basis for the decision
UNAIT classifies automatically and sets out the specific obligations
The platform

Every function. One platform.

From the first record to audit-ready evidence, UNAIT supports the whole AI governance process in your organisation.

01

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.

app.unait.de / ai-inventory
HS
AI Inventory
28 AI systems recorded
+ New case
Create audit report ▾
1 Prohibited 4 High-risk 8 Limited 15 Minimal
ChatGPT use in customer service
High-risk ✓ Approved
›
AI-assisted candidate screening
High-risk Draft
›
Predictive maintenance – plant 3
Limited ✓ Approved
›
Text correction with Microsoft 365
Minimal ✓ Approved
›
02

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.

app.unait.de / assessment
HS
Sections
✓
1.
General
›
2.
Context of use
3.
Risk review
4.
Affected persons
5.
Technical details
6.
Accountability
Step 2 of 6 · Context of use
2.3: What is the AI system primarily used for?
Select the category that matches the intended purpose of your AI system.
Supporting human decisions
Automated decision-making
Internal process optimisation
Customer communication / external use
← Back
Next →
03

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.

app.unait.de / cases / 124
HS
←
AI-assisted candidate screening
Draft
Approve
Risk classification
⚠
High-risk AI system
Annex III, point 4 EU AI Act
The system falls under Annex III of the EU AI Act (employment decisions) and therefore counts as high-risk. Extensive documentation, transparency and governance obligations apply.
Show obligation packages
Revise assessment
Export PDF
04

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.

app.unait.de / ai-inventory
HS
AI Inventory
Create audit report ▾
ChatGPT use in customer service
High-risk
AI-assisted candidate screening
High-risk
Create audit report
PDF export of all assessed AI systems
Reporting period
📅 1 Jan 2026
📅 31 May 2026
28
Cases in total
4
High-risk
19
Approved
⬇ Download PDF
Cancel
05

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 →
app.unait.de / governance-quickcheck
HS
Governance Quick-Check
Result from 28 May 2026
Overall score: 68%
AI inventory & documentation
85 %
OK
AI literacy (Article 4)
65 %
Medium
Risk assessment & monitoring
72 %
Medium
Approval processes & governance
40 %
High
Data protection & GDPR compliance
90 %
OK
Recommended actions →
06 New

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.

Processed in Germany No permanent storage AI-assisted
app.unait.de / assessment
HS
AI
AI assistant
Assessment guidance · step 2 of 6
Anonymous
Current question: 2.3: Primary intended purpose
Let me help with question 2.3. „Automated decision-making“ means the system decides without human review. Does that apply to your system?
No, a manager always reviews the result.
Thank you. Human oversight is a key criterion of the EU AI Act. I have noted this for the assessment. ✓
Your question or answer …
07 New

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.

41 document templates Role and risk based Audit-ready documentation
app.unait.de / cases / 124 / documents
HS
Compliance documents
ChatGPT use in customer service · High-risk
Export ⬇
6 of 8 relevant documents complete 75 %
✓
Risk management plan
Complete
✓
Technical documentation
Complete
✓
Human oversight plan
Complete
···
EU declaration of conformity (Article 47)
In progress
○
Registration in the EU database (Article 49)
Open
+ 3 further relevant documents
We set out to build something that supports organisations with the review obligations of the EU AI Act in just a few clicks.
Hendrik Schneider, Founder & CEO, UNAIT

AI governance.
Structured. Documented.
Ready to use.

Why UNAIT

Steer AI in a structured way instead of resolving risks after the fact

Anyone intending to meet the obligations of the EU AI Act first needs to know where AI is in use and which risks exist. Without systematic records, assessment and documentation, obligations and risks go unnoticed. UNAIT makes exactly that practical and cost-efficient. It grew directly out of the day-to-day work of a public institution.

Legal consequences

Most organisations act only once an authority asks or an incident becomes public. By then the documentation is missing, roles are unclear, and evidence can barely be reconstructed in full. A governance structure put in place early prevents exactly that.

€35 million Maximum fine for certain infringements of the EU AI Act
What UNAIT does
Risk classification: automated grading against the categories of the EU AI Act
Audit-ready documentation: traceable evidence at the touch of a button, ready for review at any time
A central AI inventory: every system in one place, clearly laid out and always current
Governance processes: clear approval paths and responsibilities
Data security: 100% hosted in Germany, GDPR-compliant
Human oversight

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.

The same answers always lead to the same classification
Every classification can be traced question by question back to the legal text
The AI assistant explains questions, it does not answer them

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

Our solutions

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.

Entry level

UNAIT Risk Tool

Ideal for organisations that want a first overview and wish to record and document individual AI systems in a structured way.

Structured assessment of individual AI applications
Risk classification under the EU AI Act
Documentation support
Orientation for staff
A documented initial assessment
Find out more
Full compliance

UNAIT Compliance

Everything in the Risk Tool plus complete AI governance: a central inventory, approval processes and audit-ready documentation.

Around 90% less time spent by AI officers
All Risk Tool functions included
A central AI inventory
Review and approval processes
Audit-ready documentation
Governance Quick-Check
Find out more
Optional add-on
Training obligation (Article 4)

KI-Wisskom training

AI literacy training via our partner KI-Wisskom, tailored to managers, staff and administration.

Role-specific learning paths
Structured e-learning modules with a certificate of completion
Designed for the requirements of Article 4 of the EU AI Act
GDPR-compliant processing
Also bookable via Secova sam®
Partner offering · ki-wisskom.training
See the training
A simple start

A governance structure in four steps

No complex IT project. UNAIT is web-based and productive within days rather than months.

01

Demo & consultation

In a 30-minute online meeting we show you UNAIT and discuss what you specifically need.

02

30 days free

Start with a free trial account, no credit card and no obligation.

03

Onboarding

We support onboarding and roll-out in your organisation personally.

04

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.