Artificial intelligence (AI): Services for safe AI

AI Compliance

Optimally prepared from a regulatory perspective: With a view to the EU AI Act and ISO standards that refer to the use of AI systems, we offer you various services.

Needs assessment

Introduction to the requirements of the EU AI Act
Evaluation of applicability
Identification of necessary measures

Risk classification

Classification of your AI application according to the EU AI Act
Determination of the required testing level
Optional: Creation of a roadmap for achieving compliance

Readiness analysis

Identification of compliance gaps & risk areas
Gap analysis of an Artificial Intelligence Management System (AIMS)
Evaluation based on a use case

Internal assessments

Support in the development or updating of internal audits & guidelines for the use of AI
Creation of an implementation guide

In-depth testing

Review of currently deployed AI systems & practices
Identification of compliance gaps & risk areas
Delivery of the results in the form of a detailed test report

Trainings

Trainings and workshops with regard to the requirements of the EU AI Act, ethical AI development and best practices for compliance

AI Trainings

We provide you with the best possible support on your way to a secure and reliable AI solution with training courses tailored to your individual needs.

Basics

Would you like to expand your knowledge regarding the basics of AI? No problem!

Our AI training courses will introduce you to the world of artificial intelligence, shed light on possible areas of application and the potential, but also the challenges and limitations of this new technology. You will learn more about how it works and the basic principles and receive valuable ideas on how you can use AI in your company.

Topic-specific

In addition to the basics, we also offer trainings with an individual, specific focus.

Possible are for example:

Secure or robust AI development
AI risks & vulnerabilities
Robustness risks & requirements
AI-specific data protection requirements

AI Red-Teaming

Putting your AI to the test: We subject your AI system to an acid test and use pentests to uncover existing security gaps and vulnerabilities - before others do.

Needs assessment

Introduction to the requirements of the EU AI Act
Identification of necessary measures
Development of further steps

Risk analysis

Evaluation of your AI application for classification under the AI Act
Generation of a risk assessment report
Optional: Development of a roadmap for compliance based on risk levels

Trainings

Training and workshops on the risks and vulnerabilities of AI, secure development and best practices

Pentesting according to OWASP LLM TOP 10

Execution of technical pentests
Comparison with the requirements according to OWASP Top 10 for Large Language Model Applications
Receipt of a detailed test report
Optional: Recommendations for action to achieve compliance

Pentesting according to OWASP ML TOP 10

Execution of technical pentests
Comparison with the requirements according to OWASP Top 10 Machine Learning Security Risks
Receipt of a detailed test report
Optional: Recommendations for action to achieve compliance

AI Robustness Check

Bye bye robustness issues: our experts assess potential risks and comprehensively check the technical robustness of your AI application.

Needs assessment

Implementation of a scoping workshop
Identification of necessary measures
Definition of requirements
Development of further steps

Risk analysis

Evaluation of the robustness risks of your AI application
Generation of a risk assessment report
Optional: Development of a roadmap for compliance based on risk levels

Testing according to catalog

Testing according to our Trusted Product Scheme (including technical in-depth testing)
Testing according to existing regulations & adaptation of AI requirements
Development of an implementation guide

In-depth testing

Technical in-depth check of the robustness of your AI system
Optional: Recommendations for achieving compliance

Training

Trainings or workshops on relevant robustness risks/requirements, robust AI development and best practices

AI Privacy Check

Our AI privacy check expands the traditional GDPR compliance assessment to include AI-specific privacy concerns.

Needs assessment

Identification of necessary measure
Development of further steps

Reconstruction of sensitive data

Technical reconstruction of sensitive data from AI-based pseudo-anonymization methods
Technical reconstruction of sensitive data from the training data of AI systems
Final evaluation by experienced data privacy experts

Proof of "right to be forgotten"

Providing technical proof that personal data has been "forgotten" by the AI model

Risk impact assessment

Evaluation of the (residual) risk of the entire system

GDPR & AI expert opinions

Assessment of GDPR and AI-specific requirements (e.g. robustness, red teaming, AI Act)