Artificial Intelligence Assessment

  • Background:

    Artificial Intelligence (AI) offers wide-ranging possibilities for society, but it entails potential risks as well. For instance, discriminatory AI reportedly occurred in automated applicant selection processes and medical diagnostic procedures. Various initiatives emerged to define common standards for AI assessment and mitigate risks such as discriminatory AI. However, defining suitable criteria and operationalizing AI assessment remains challenging. As a result, up until now, there are no common AI assessment standards. Hence, we aim to contribute to a better understanding of AI assessment challenges.


    Illustrative Thesis Objectives:

    • Examining AI assessment challenges (either based on general AI characteristics or highlighting specific aspects such as explainability, fairness, etc.) (method: e.g., literature review)
    • Validating AI assessment challenges in practice (method: e.g., qualitative interview study / expert interviews)
    • Thesis projects related to current AI assessment initiatives (e.g., certification, quality marks, codes of conduct, codes of ethics, ethics principles, etc.)
    • Present your thesis project idea related to AI Assessment


    Introductory Literature:

    • F. T. Jaigirdar, C. Rudolph, G. Oliver, D. Watts and C. Bain, "What Information is Required for Ex-plainable AI? : A Provenance-based Research Agenda and Future Challenges," 2020 IEEE 6th International Conference on Collaboration and Internet Computing (CIC), 2020, pp. 177-183, doi: 10.1109/CIC50333.2020.00030
    • P. Cihon, M. J. Kleinaltenkamp, J. Schuett and S. D. Baum, "AI Certification: Advancing Ethical Practice by Reducing Information Asymmetries," in IEEE Transactions on Technology and Society, vol. 2, no. 4, pp. 200-209, Dec. 2021, doi: 10.1109/TTS.2021.3077595
    • German Standardization Roadmap on Artificial Intelligence. DIN Deutsches Institut für Normung e.V.; DKE German Commission forElectrical, Electronic & Information Technologies of DIN and VDE. -