Jump to the main content

TRUSTED AI by TÜV AUSTRIA

A Member in Good Standing of TIC Council | TÜV®

TRUSTED AI by TÜV AUSTRIA

Overview

Artificial intelligence put to the test

The world's first machine learning certification scheme

TRUSTED AI by TÜV AUSTRIA is the world's first Machine Learning (ML) certification scheme. We have deliberately chosen to start certification with supervised learning models, as these are already used in many technological applications in industry. They are characterized by clearly defined problem definitions and allow a formal interpretation and validation.

The certificate TRUSTED AI by TÜV AUSTRIA confirms robustness, safety and suitability of a certified AI application for defined purposes and fields of application. Currently, our certification is still limited to supervised learning models, other ML models will follow successively.

Request certificationWhitepaper Trusted AI

TRUSTED AI by TÜV AUSTRIA Certificate

Security for developers and operators

TRUSTED AI by TÜV AUSTRIA independently confirms to AI application developers the correctness and suitability of their AI application for the specified purposes.

Independent certification by TÜV AUSTRIA provides developers and users of AI application, assurance about the quality, security and suitability of the certifies application. This also offers further assurance to investors, for example when selecting suitable development partners and/or applications.

Request certification

Request AI application for certification

Security in software development

21 Audit categories for protection against external tampering

  • Risk assessment of application
  • Technical stability and safety
  • Secured development environments
  • Software security tests
  • Secure use of cloud technologies
  • Handover and monitoring in operation at the customer
  • and many more...

Functional requirements

12 Audit categories for data, methodology, model selection & documentation.

  • Data: Acquisition, preprocessing, analysis
  • Model: selection, requirements, qualitative inspection and deployment
  • Error handling
  • Documentation and communication
  • Explainability and interpretability
  • and many more...

Processing of personal information

9 Audit categories for ethics requirements

  • Fundamental rights
  • Priority of human action
  • Avoidance of objectively unjustified discrimination
  • Accessibility (barrier-free)
  • Consideration of stakeholders
  • and many more...

Data protection

14 Audit categories for the protection of data

  • Data protection by design oder by default
  • Data minimization
  • Rights of the data subjects
  • Automated decisions in individual cases including profiling
  • Audits on commissioned data processing
  • Data breach detection
  • and many more...

TRUSTED AI by TÜV AUSTRIA Certification procedure

The TRUSTED AI by TÜV AUSTRIA certification process takes three - four months from commissioning to certification.

  • Catalog of requirements

    • Testing
      • Functional requirements and ethics and data protection

  • Audit

    Our audits consist of document reviews, interviews, and technical inspections of software development processes and functional requirements as well as information security checks. We check both the appropriateness of the specifications for processes, procedures, and methods used, as well as their implementation and effectiveness.

    • Documentation
      • Process description
      • AI model specifications
      • Data processing
    • Interviews
      • Implementation documented process

  • Technical inspection

    • Machine Learning Artifacts
      • Test reports
      • Monitoring information
      • Measurement results
    • Protection against tampering

  • Certificate

  • Annual monitoring audit

  • Re-certification after 3 years

Contact us!

  •  | Print
to top