Kloepfel Group Comments On Position Paper By 14 Leading Associations
The Kloepfel Group, specializing in supply chains and procurement, welcomes the adoption of the AI regulation by the European Union on May 21, 2024, by all 27 EU member states. Marc Kloepfel, CEO of the Kloepfel Group, stated, “The regulations are important for ensuring the safety and trust in artificial intelligence (AI). However, overly strict regulations should not stifle innovation.”
The Kloepfel Group examined the position paper of 14 leading German business associations on the implementation of the AI regulation from the perspective of supply chain management using concrete examples.
1. Precise Definition of Artificial Intelligence
The definition of AI must be precise to avoid unnecessary bureaucracy. Many systems in supply chains are data-based and analytical, but not true AI. Therefore, conventional IT systems should not fall under the AI definition.
Example: An inventory management system that collects data and generates reports should not be classified as AI since it does not use machine learning or self-optimizing algorithms. These systems analyze data but do not autonomously generate new insights or optimizations.
2. Risk Assessment
It is important to consider the specific purpose of an AI application and the associated risks. For supply chain applications that are primarily supportive, the regulation should be pragmatic and innovation-friendly.
Example: An AI tool that creates supplier ratings based on historical performance data should be classified as less risky compared to an autonomous vehicle for goods delivery. The evaluation tool merely supports decisions and carries a lower direct risk.
3. Avoiding Double Regulation
New requirements under the AI regulation must be consistent with existing regulations, such as the GDPR, to avoid additional bureaucratic hurdles.
Example: An AI-supported system for analyzing supply chain risks that processes personal data must already comply with the General Data Protection Regulation (GDPR), including conducting a Data Protection Impact Assessment (DPIA). Instead of performing double evaluations, the existing DPIA should be used to demonstrate compliance with both regulations.
4. Uniform EU Rules
A uniform implementation of the new requirements across the entire EU is essential to prevent a “patchwork” of regulations. This is particularly critical for internationally operating supply chains.
Example: A multinational company using AI to optimize supply chain processes should not face different requirements in every EU country. Uniform rules would avoid efficiency losses and high costs.
5. Effective Administrative Structure
Creating a well-functioning administrative structure with knowledgeable personnel is crucial. Duplicate structures should be avoided, particularly in already heavily regulated industries.
Example: For AI applications in procurement and supply chain management, a central authority should be responsible, possessing both AI expertise and knowledge of the respective industry. This would avoid redundancies and facilitate compliance with the regulations.
6. European Collaboration and Cooperation
The AI Office, established by the European Commission as a center of AI expertise across the EU, is particularly important. Adequate staffing and close collaboration with businesses, institutions, experts, and stakeholders are key. The AI Board, a central body of European governance, consists of representatives from member states. Its tasks include coordinating national supervisory authorities and providing expert knowledge and best practices.
Example: For supply chains, it is crucial that the AI Office and AI Board work closely with logistics companies and technology providers. Regular workshops and exchange formats can contribute to developing and implementing practical solutions for the use of AI in supply chains.
7. Practical Guidelines for Implementation
Practical and understandable guidelines are necessary to help developers of AI applications comply with the regulations. These guidelines could be developed by the European Commission, national authorities, and representatives from business and civil society.
Example: A checklist with step-by-step instructions for evaluating and documenting AI applications in supply chain management could help developers effectively implement the regulations.
8. Protection of Intellectual Property
The protection of intellectual property and copyrights is essential, especially in the development of generative AI, which is used in many supply chain processes.
Example: A company using generative AI to optimize packaging designs should ensure that the original images and data used are licensed to avoid copyright violations.
Conclusion:
Efe Duran Sarikaya, CEO of Kloepfel Consulting, concluded: “We support the goals of the AI regulation. A pragmatic, unified, and innovation-friendly regulation is necessary to fully harness the benefits of AI while strengthening supply chains. Only in this way can we fully realize the potential of this key technology for growth, productivity, and employment.”
Contact:
Kloepfel Group
Damir Berberovic
Tel.: 0211 941 984 33 | Mail: rendite@kloepfel-consulting.com