Our Approach to
Artificial Intelligence
As an education provider for Artificial Intelligence, we practice what we teach: Transparent, responsible, and ethical AI usage. This statement fully documents where and how we use AI.
AI Usage on Our Website
This website was significantly developed with the support of AI systems. We transparently disclose which areas were created with AI assistance.
| Area | AI System | AI Assistance | Quality Assurance |
|---|---|---|---|
| Website Structure | Claude AI (Anthropic) | Substantial | Human Design & Testing |
| HTML/CSS Code | Claude AI | Heavily Assisted | Code Review & Adjustments |
| JavaScript Functions | Claude AI | Assisted | Testing & Debugging by Team |
| Text Content | Claude AI / GPT-4 | Partial | Expert Review & Adaptation |
| Design Concept | Claude AI | Contributing | Human Design Decisions |
| Visual Elements | Various (see Section 02) | Variable | Selection & Editing by Team |
| AI Radar Newsletter | Claude AI (Anthropic) | Substantial | Editorial Review & Approval |
CIRA - Our AI Chat Assistant
On our website, CIRA (CITECH Intelligent Response Assistant) is available to you as a chat assistant.
Technical Foundation
Claude API (Anthropic) with RAG system (Retrieval Augmented Generation)
How It Works
CIRA searches our website content and generates responses based on current information. All responses are AI-generated.
Data Privacy
Your queries are processed via the Claude API. No conversation data is used for training purposes (Zero Data Retention Policy).
AI-Generated Images and Graphics
We use AI tools for creating and editing visual content.
| Element | Tool Used | Status |
|---|---|---|
| Profile Image Marcel Batur | AI Image Generator | Fully AI-generated |
| Website Animations | CSS/JavaScript (Claude AI) | AI-programmed, Human-tested |
| Icons & Symbols | Standard Libraries / Unicode | Not AI-generated |
| Graphic Elements | Variable (AI tools as needed) | Varies by element |
AI as a Teaching Tool, Not the Teacher
In our training courses, we use AI systems exclusively as demonstration objects and practice tools. All content is delivered by certified instructors.
Systems Used
| System | Purpose | Data Privacy |
|---|---|---|
| Claude / ChatGPT | Text generation, Prompt Engineering, Code assistance | Cloud-based, no data storage |
| Stable Diffusion | Image generation (local) | 100% local, no cloud |
| Ollama | Local LLMs | 100% offline, full data control |
| LangChain | RAG systems, Chains | Framework, privacy depends on setup |
| Python ML Libraries | Scikit-learn, TensorFlow, PyTorch | 100% local |
Training Content
Our training materials (presentations, handouts, exercises) are partially created with AI support and subsequently undergo comprehensive expert review:
- Presentation slides: AI-assisted, fully reviewed by experts and didactically adapted
- Exercises: Partially AI-generated, pedagogically validated and aligned with learning objectives
- Code examples: Developed with AI assistance, all tested and documented
- Handouts: AI-assisted creation, validated by experts and kept up to date
VERIFY Process for AI Content
Every AI-generated piece of content goes through our structured quality process.
V — Validate
Expert review by qualified specialists. All AI outputs are checked for accuracy, relevance, and quality.
E — Edit
Adaptation to target audience, corporate language, and didactic requirements. Removal of errors or inaccuracies.
R — Review
Second review by an independent reviewer. Four-eyes principle for critical content.
I — Integrate
Integration into existing structures with consideration for consistency and corporate identity.
F — Finalize
Final approval by management. Last quality check before publication.
Y — Yield
Publication with clear AI labeling and continuous quality monitoring in live operation.
Regulatory Classification
According to the EU AI Act, our AI systems fall into the following categories.
| System | Risk Class | Compliance Measures |
|---|---|---|
| Website (Claude AI) | Minimal Risk | Voluntary transparency statement (this page) |
| CIRA Chatbot | Limited Risk | Transparency obligation fulfilled, labeling in place |
| Training AI | Minimal Risk | Demonstration purpose, no critical application |
| AI Radar Newsletter | Minimal Risk | AI-assisted content creation, editorial final review, transparency notice in newsletter |
Transparency & Accountability
You have the right to full information about our use of AI.
Information
Upon request, we will inform you whether and how AI was used in the creation of specific content.
Labeling
We label AI-generated content wherever possible and appropriate.
Quality Guarantee
In the event of errors in AI-generated content, we take full responsibility and correct them promptly.
Data Privacy
Your data is processed in accordance with GDPR, including during AI interactions.
Feedback
Report quality issues or concerns regarding AI content to us.
Responsible AI Usage
We are aware of the environmental impact of AI systems and make conscious decisions to conserve resources.
Energy Efficiency
Preference for smaller, more efficient models where possible. Use of local systems (Ollama, Stable Diffusion) reduces data center load.
Resource Conservation
Avoidance of unnecessary AI requests. Caching of results. Deliberate use instead of automated mass production.
Education & Awareness
Training our clients in energy-efficient AI usage. Addressing sustainability aspects in all courses.
Local Alternatives
Where possible, we use local AI systems (Ollama, Local LLMs) instead of cloud services. This reduces data transfer and server load.
Creative Collaboration
Transparency also means making our partners visible.
- Corporate Design & Web Design: B_Line Graphix — Development of the visual identity, logo design, and creative conception of the website
This collaboration does not involve AI usage on the part of the partners in the design area.
Pioneers for AI Transparency
We continuously work on setting standards for AI transparency.
- Industry Standard: Development of an AI Transparency seal for education providers
- Open Source: Sharing our findings and best practices with the community
- Research: Collaboration with universities on AI ethics and transparency
- Education Initiative: Free workshops on responsible AI usage
- Continuous Improvement: Quarterly review and update of this statement
Transparency is not a competitive disadvantage, but the foundation for trust in the AI era. We lead the way.
AI Transparency Officer
Marcel Batur
Managing Director & AI Officer
For questions about AI transparency, information requests, or feedback on AI-generated content.
m.batur@citech-ai.deVersion History
| Date | Version | Change |
|---|---|---|
| 17.02.2026 | 2.1 | Addition: AI Radar Newsletter as AI system in Sections 01 and 07 |
| 27.10.2025 | 2.0 | Complete redesign: Structured documentation per compliance standard, full disclosure of all AI systems, VERIFY process added |
| 27.09.2025 | 1.1 | Addition: CIRA Chat Assistant with Claude API and RAG system |
| 01.09.2025 | 1.0 | Initial publication of the AI Transparency Statement |
Last Review: February 2026 | Next Review: May 2026