AI governance · EU AI Act

AI legally binding use

We make your AI use compliant: risk classification, transparency obligations and documentation – implemented pragmatically – in time for the obligations from 2 August 2026.

Book a consultationView services
Risk classificationTransparency obligationsAudit-ready
EU AI Act ReadinessYour compliance status
Creating an AI inventory
Risk classes
Transparency obligations implemented
Documentation verifiable
Transparency and prohibitionsfrom 02.08.2026
Deadline: 2 August 2026
From this date, the prohibited AI practices (Art. 5) and the transparency and labelling obligations (Art. 50) of the EU AI Regulation apply. The requirements for high-risk systems were combined with the Digital Omnibus postponed to 2 December 2027 (Annex III) or 2 August 2028 (Annex I). Postponed does not mean done: risk classification, logging and documentation are architectural decisions that are now being made.

The EU AI Act also affects you

Anyone who uses AI must classify risks, create transparency and document them. Without preparation, there are fines and liability risks.

  • Unclear obligations: Which AI application falls into which risk class?
  • The deadline of 02.08.2026 is approaching – without preparation it will be tight.
  • Transparency obligations: AI content and chatbots must be labeled.
  • Lack of documentation and governance become a liability risk.

The four risk classes of the EU AI Act

The EU AI Act ranks AI systems according to their risk – the higher the class, the stricter the obligations. We arrange your applications safely and derive the right measures.

1
Unacceptable riske.g. social scoring, manipulative or exploitative systems
Prohibitions
2
High riske.g. AI in recruiting, creditworthiness, medicine, critical infrastructure
Strict obligations
3
Limited riske.g. chatbots & AI-generated content – must be labeled
Transparency obligation
4
Minimum riske.g. spam filters, AI in games, simple recommendations
No special requirements

Simplified presentation of the four risk classes of the EU AI Act. We carry out the specific classification of your applications together.

Our governance services

From inventory to risk classification to audit-proof documentation.

AI Inventory & Risk Class

We collect your AI applications and assign them to the risk classes of the AI Act.

Transparency and labelling

Implementation of transparency obligations – from chatbot labelling to AI content.

Documentation and Processes

Governance policies, evidence and processes that stand up to scrutiny.

This is how we proceed

Pragmatic from inventory to ongoing monitoring.

1. Inventory

Capture and evaluate all AI applications.

2. Risk class

Classification under the EU AI Act and derivation of duties.

3. Measures

Implement transparency, technical and organizational measures.

4. Documentation

Create proof and guidelines in an audit-proof manner.

5. Monitoring

Ongoing monitoring and adaptation in the event of changes.

Why act now

2026
AI Act date 02.08. – prepared in time
4
Risk classes clearly identified
100%
documented & audited secure

Key figures from real ALGEBRA projects and studies. Concrete results depend on the use case.

What our AI governance covers

  • AI inventory and use case capture
  • Risk classification according to the EU AI Act
  • Transparency and labelling obligations (chatbots, AI content)
  • Technical and organizational measures
  • Governance guidelines and audit-proof documentation
  • Ongoing monitoring and adaptation to changes

Related benefits

AI Consulting

RAG systems & AI agents from your knowledge.
Find out more

Individual software

Customized applications for your processes.
Find out more

Business Intelligence & ETL

From data reliable key figures and dashboards.
Find out more

Frequent questions about the EU AI Act

The regulation is staggered. As of 2 August 2026, the prohibitions (Art. 5) as well as the transparency and labelling obligations (Art. 50) apply; for generative systems that are already on the market, a transitional period applies until 2. December 2026. The obligations for high-risk systems were postponed by the Digital Omnibus: to 2 December 2027 for stand-alone systems (Annex III) and to 2 August 2028 for AI as a security component in products (Annex I). The relevant publication is in the Official Journal of the EU.

Yes – as soon as you use or offer AI. The scope and obligations depend on the risk class of your applications.

For example, users need to recognize when they interact with an AI system (chatbot) or when content is AI-generated.

No. We create the necessary documentation and governance with you – pragmatic and audit-proof.

We start with an AI inventory, classify the risk classes and derive concrete measures.

Make your AI use legally secure

The deadline of 02.08.2026 is coming – and the high-risk obligations follow 2027/2028. Start early enough: with an AI inventory and a clear action plan.

Booking an appointment now