Artificial Intelligence in German employment relationships: Analysis from an international perspective

RA Daniel Wasser LL.M.; Prof. Dr. Vagelis Papakonstantinou

doi:10.5771/9783748971948

Summary

Artificial intelligence is increasingly used in employment relationships, yet many employers remain hesitant. One reason is the lack of a shared technical and legal understanding of what constitutes AI; another is the wide divergence of legal requirements across jurisdictions, including within the EU. Companies operating internationally therefore often refrain from introducing AI in the employment relationships of their foreign subsidiaries. Against this background, this handbook focuses on the implementation of AI in German employment relationships and seeks to address and reduce common reservations about its use in the workplace.

Keywords

Search publication

Bibliographic data

Edition: 1/2026
Copyright Year: 2026
ISBN-Print: 978-3-7560-4333-0
ISBN-Online: 978-3-7489-7194-8
Publisher: Nomos, Baden-Baden
Series: The Regulation of Digital Technologies
Volume: 3
Language: German
Pages: 335
Product Type: Edited Book

Titelei/Inhaltsverzeichnis No access Pages 1 - 6
Introduction No access Pages 7 - 8
1. Authors:
  1. I. Preliminary Note No access
  2. II. Introduction No access
  3. III. Definitions No access
  4. IV. Targeted, measurable, and appropriate deployment No access
  5. V. Expert Systems No access
  6. VI. Linear Models No access
  7. VII. Deep Learning No access
  8. VIII. Neural networks No access
  9. 1. 1. Large Language Models (LLMs) and statistics No access
    2. 2. Multimodal models No access
    3. 3. Agentic systems No access
    4. 4. Knowledge structures and data integration No access
    5. 5. RAG (Retrieval-Augmented Generation) No access
    6. 6. Building a database No access
    7. 7. Retrieval No access
    8. 8. Augmentation & generation No access
    9. 9. Ontologies No access
    10. 10. Knowledge Graphs No access
  10. X. Conclusion No access
  11. Reference list No access
2. Authors:
  1. 1. I. Introduction No access
    2. 1. a. Defining Discrimination No access
      2. 1. Flaws in Algorithmic Design No access
        2. Bias in Data No access
        3. Intransparency (Black Box) No access
        4. Self-Reinforcing Discrimination through Feedback Loops No access
    3. 1. ii. Data Protection No access
    4. 1. 1. German Constitution (Grundgesetz) No access
      2. 2. General Equal Treatment Act (AGG) No access
      3. 2. German Federal Data Protection Act (BDSG) No access
      4. 3. German Works Constitution Act (BetrVG) No access
      5. 4. Impact of discrimination on employees throughout the employment relationship No access
      6. a. Recruitment / Initial Stage of Employment No access
    5. 1. 1. Enhance Employee Training and AI Literacy No access
      2. 2. Ensure Transparency and thorough documentation: No access
      3. 3. Develop clear internal policies and contracts with providers No access
      4. 4. Use diverse and representative training Data No access
      5. 5. Establish comprehensive Algorithmic Impact Assessments (AIAs) No access
      6. 6. Maintain Human Oversight (Human-in-the-Loop) No access
      7. 7. Integrate Worker Participation No access
      8. 8. Establish Accessible Complaint and Regress Mechanisms No access
      9. 9. Collaborate in Ethical Ecosystems No access
    6. VI. Challenges and Outlook No access
    7. Reference List No access
3. Authors:
  1. I. Developing the understanding of data protection No access
  2. II. 1950–1980: Solely few ethical guidelines and start of the ‘first AI-Winter’ No access
  3. III. Still mainly non-AI specific Regulations from 1980–2000 No access
  4. 1. 1. Non-governmental initiatives No access
    2. 2. Guidelines and other instruments from or in cooperation with the European Union No access
    3. 3. Legislative Actions from the European Union No access
    4. 4. Peculiarities in Germany No access
    5. 5. Latest Developments in Legislation Regarding AI-generated Deepfakes No access
  5. V. Conclusion No access
  6. Reference List No access
4. Authors:
  1. 1. 1. Problem definition and relevance No access
    2. 2. Objectives and scope of the study No access
    3. 3. Course of the investigation No access
  2. 1. 1. Automated decisions and liability principles No access
    2. 1. a. Fundamentals of labor law No access
      2. b. Contractual and tortious liability No access
      3. c. Product liability No access
      4. d. EU AI Regulation No access
      5. e. General Data Protection Regulation No access
      6. f. Interim result No access
  3. 1. 1. 2. Employee liability for incorrect use of AI No access
      2. a. Incorrect operation of an AI system No access
      3. b. Insufficient review of AI decisions No access
      4. c. Expected monitoring obligations of the employee No access
      5. d. Disregard of operational guidelines or safety mechanisms No access
      6. e. Liability for AI as vicarious agents No access
    2. 1. a. The role of employee will in autonomous AI decisions No access
      2. b. Contributory negligence of the employer No access
    3. 4. Causality problems No access
    4. 5. Case law and statutory regulations No access
    5. 6. Interim result No access
  4. 1. 1. a. Liability for operational risks No access
      2. b. Liability according to § 278 BGB for vicarious agents No access
      3. c. Liability under Section 831 BGB for AI as vicarious agent No access
    2. 1. a. Implementation and selection of suitable AI systems No access
      2. b. Training and instruction of employees No access
      3. c. Establishment of control and monitoring mechanisms No access
    3. 3. Data protection obligations No access
    4. 4. Interim result No access
  5. V. Manufacturer and developer liability No access
  6. VI. Need for reform and influence of the AI Regulation No access
  7. VII. Conclusion No access
  8. Reference list No access
5. Authors:
  1. I. Introduction No access
  2. 1. 1. Essential cybersecurity principles No access
    2. 2. Intersection of AI and cybersecurity No access
    3. 3. Unique risks in employment-related AI systems No access
    4. 4. Practical example: recruitment platforms No access
  3. 1. 1. The General Data Protection Regulation (GDPR) No access
    2. 2. The German Federal Data Protection Act (BDSG) No access
    3. 3. The EU AI Act Proposal No access
    4. 4. Practical Example: Compliance Adaptation by German Companies No access
    5. 5. Concluding Observations No access
  4. 1. Data poisoning: Manipulation of training data No access
    2. Adversarial attacks: Exploiting model weaknesses No access
    3. Privacy breaches: Exposure of sensitive employee data No access
    4. Practical example: Vulnerabilities in AI monitoring software No access
    5. Conclusion No access
  5. 1. Technical safeguards No access
    2. Organisational measures No access
    3. Employee training and awareness No access
    4. Practical example: AI-driven performance monitoring tools No access
  6. 1. Employer liability No access
    2. Employee rights and obligations No access
    3. Case study analysis No access
    4. Regulatory and judicial responses No access
    5. Practical example No access
    6. Outlook No access
  7. 1. Divergences between Germany and the United States No access
    2. The United Kingdom as a middle path No access
    3. Emerging jurisdictions and global fragmentation No access
    4. Practical compliance challenges No access
    5. Case examples of multinational strategies No access
    6. Recommendations for harmonisation No access
  8. VIII. Conclusion No access
  9. Reference list No access
6. Authors:
  1. I. Artificial Intelligence in the United Arab Emirates No access
  2. 1. 1. Regulations directly governing the use of AI No access
    2. 2. Regulations indirectly affecting and governing the use of AI No access
  3. 1. 1. a. Lawful Basis for Data Processing No access
      2. b. Automated processing and Individual decision-making No access
      3. c. Requirements of non-discrimination No access
    2. 1. a. Principle of duty to personal work performance No access
      2. b. Declarations of intent under UAE labour law: disciplinary actions, notices, and terminations No access
      3. c. Data protection aspects No access
      4. d. Protection of trade secrets and confidential information in the context of AI-system use No access
  4. IV. Conclusion No access
  5. Reference list No access
1. Authors:
  1. A. Introduction No access
  2. B. The EU Directives 2019/770 and 2019/771 for the supply of digital content, digital services and goods with digital elements No access
  3. 1. C.1. Preliminary step: Definition of AI according to the existing EU legal framework No access
    2. C.2. Intermediate step: Software as ‘digital content’ or ‘digital service’ or ‘good with digital elements’ under DSC and SCD No access
    3. C.3. Interim conclusion: AI as ‘digital content’, ‘digital service’ or ‘goods with digital elements’ No access
  4. 1. D.1. The special nature of AI technology No access
    2. 1. (a) Definition of contractual object: No access
      2. (aa) Conformity assessment based on subjective requirements: No access
        (bb) Conformity assessment based on objective requirements: No access
        (cc) Conformity assessment and third-party rights: No access
      3. (c) Remedies: No access
      4. (d) Modification: No access
  5. E. Conclusions No access
  6. Bibliography No access
2. Authors:
  1. 1. Introduction No access
  2. 2. Special features of individual rights No access
  3. 1. 2.2. Employer's right to give instructions and limits on the use of AI No access
    2. 2.3. AI company guidelines No access
  4. 1. a) Prohibited practices No access
    2. b) High-risk AI systems No access
  5. 1. 3.1. Consultation of experts No access
    2. 3.2. Duty to inform No access
    3. 3.3. Co-determination rights No access
3. Authors:
  1. A. Introduction No access
  2. 1. B.1. Operation and traits of autonomous AI systems No access
    2. B.2. AI in recruitment No access
  3. 1. 1. C.1.1. No access
      2. C.1.2. No access
      3. C.1.3. No access
      4. C.1.4. No access
    2. 1. C.2.1. No access
      2. C.2.2. No access
      3. C.2.3. No access
    3. 1. C.3.1. No access
      2. C.3.2. No access
  4. D. Civil liability for unequal and discriminative treatment during AI recruitment No access
  5. E. Concluding remarks No access
  6. Bibliography No access
4. Authors:
  1. I. Introduction and examples of AI-based behaviour control No access
  2. 1. 1. Classification as a high-risk AI system No access
    2. 2. Regulatory requirements for high-risk AI systems No access
  3. 1. 1. a) Special labour law regulation of Art. 88 GDPR in conjunction with § Section 26 (1) sentence 1 Federal Data Protection Act (Bundesdatenschutzgesetz; BDSG) No access
      2. b) Consent pursuant to Section 26 (2) BDSG or Art. 6 (1) a), 7 GDPR No access
      3. c) Collective agreement pursuant to Art. 88 GDPR in conjunction with § Section 26 para. 4 BDSG No access
      4. d) Legitimate interest pursuant to Art. 6 para. 1 f) GDPR No access
    2. 2. Rights of data subjects No access
    3. 3. Processing, Art. 28 GDPR No access
    4. 4. Joint controllership, Art. 26 GDPR and reference to third countries No access
    5. 5. Data protection impact assessment, Art. 35 No access
    6. 6. Prohibition of automated decision-making, Art. 22 GDPR No access
  4. 1. 1. Consultation of an expert pursuant to Section 80 (3) BetrVG No access
    2. 2. Right of co-determination pursuant to Section 87 (1) No. 1 BetrVG No access
    3. 3. Right of co-determination for the introduction and use of technical equipment pursuant to Section 87 (1) No. 6 BetrVG No access
    4. 4. Co-determination at AI in occupational health and safety No access
    5. 5. Planning of work procedures and processes pursuant to Section 90 (1) No. 3 (and No. 2) BetrVG No access
    6. 6. Selection guidelines pursuant to Section 95 (2a) BetrVG No access
    7. 7. Operational change pursuant to Section 111 BetrVG if AI changes work processes, operational organisation, operational purpose or operational facilities No access
  5. V. Anti-discrimination law No access
  6. 1. 1. Personal rights of employees No access
    2. 2. Equity as the limit of the right to issue instructions No access
  7. Reference list No access
5. Authors:
  1. I. Introduction No access
  2. II. General principles of creatorship on work results in employment No access
  3. III. International Perspective on Copyright in Employment Relationships No access
  4. IV. German Legal Framework for Copyright in Employment No access
  5. V. AI-generated and AI-assisted creation No access
  6. VI. Protectability of AI Generated Work Results No access
  7. VII. Impact of Non-Protection on Certain Processes and Business Models No access
  8. VIII. Labelling Requirements No access
  9. IX. Copyright/Trademark Infringements in Work Results by Unlicensed AI Training Data No access
  10. X. Liability of Employer No access
  11. XI. Practical Considerations for Employers in Germany No access
  12. XII. Possibilities of German Employment Law No access
  13. XIII. Practical Checklist for Employers No access
  14. XIV. Conclusion No access
  15. Reference list No access
6. Authors:
  1. I. Introduction No access
  2. 1. 1. Job transformation No access
    2. 1. a) Organisation of work and conduct of employees (section 87 para. 1 no. 1 BetrVG) No access
    3. 1. c) Health and safety of employees (section 87 para. 1 no. 7 BetrVG) No access
    4. 1. a) AI expert (section 80 para. 3 sentence 2 BetrVG) No access
      2. b) Information and consultation (section 90 para. 1 no. 3 and para. 2 sentence 1 BetrVG) No access
      3. c) Approval of AI-generated selection criteria (section 95 para. 2 lit. a BetrVG) No access
      4. d) Privacy protection (section 75 para 2 BetrVG) No access
      5. e) Change in operations (section 111 BetrVG) No access
  3. III: Conclusion No access
  4. Bibliography No access
Authors No access Pages 333 - 335