EU Data Strategy
The paper ”EU Data Strategy: Rethinking the Role of Ethics in Shaping the New Rules of Competition Law in the Era of Data Governance” (Montero Santos, L., Bey, N.E.H.) will be presented at the...
The paper ”Internet of Things and the Governance of Big Data and Artificial Intelligence” (Knieps, G. ) will be presented at the 11th FSR Annual Conference “From Data Spaces to Data Governance” (9-10 June, 2022).
The collection and processing of large data sets (big data) does not come in isolation but is evolving based on a combination of data collection innovations due to rapidly decreasing (camera based) sensor technologies, strongly increasing computer processing and storage capacities as well as innovations in communication technologies. In meantime, big data cloud computing is gaining increasing relevance for local, regional, and cross-country data traffic driven by the high bandwidth capacities of 5G networks. Cloud computing is no longer only a cost-saving possibility to an on-demand shared pool of computing and storage solutions but is gaining a pivotal function in the entrepreneurial design of data value chains for real-time, adaptive sensor-based applications in many use cases of the Internet of Things (IoT). Classes of use cases requiring a tactile Internet (e.g., driverless networked vehicles, smart manufacturing, augmented and virtual reality) are concomitant with massive and high-velocity data sets challenging the traditional approaches (e.g., statistical analysis or optimization theories) to derive the relevant decisions based on the insights from this data. This is the very reason why Artificial Intelligence (AI) relying on algorithms-based pattern recognition becomes so relevant within the network industries of the future. Although AI has the character of an umbrella term with different meanings, its basic principle is the interaction of sensors, operational logic (algorithms) and actuators. For a well-defined set of objectives, the operational logic of data trained algorithms depending on the sensor-based data inputs a set of actuator decisions is chosen. The purpose of this paper is to analyze the changing entrepreneurial incentives for data portability and data sharing and the changing needs for regulations to tackle the complexities of upcoming AI-powered big data virtual networks.
Research Design and Expected Results:
An important precondition for the development of AI is the access to big data. The focus is therefore on two topical complementary waves of reform initiatives, the “EU data strategy” on one hand and the “AI for Europe strategy” on the other hand. Based on the “European Data Economy”, legal framework the recently established self-regulatory “SWIPO (switching cloud providers and porting data) Codes of Conduct” on non-personal data portability are analyzed focusing on the reduction of risk of vendor lock-in by cloud service providers as well as cloud security certifications. The “GAIA-X project” for the next generation of a data infrastructure for Europe is analyzed with particular focus on the role of codes of conduct between European and non-European cloud service providers.
To analyze the potentials of AI in network industries the analytical concept of AI-powered big data virtual networks is elaborated. Although several actors may be involved such as broadband providers, cloud service providers, geopositioning-service providers, or sensor network service providers, the final responsibility for bundling these different service components lies in the hand of the AI-powered big data virtual network providers. Platform operators responsible for the performance guarantees on the physical side of IoT applications may be horizontally integrated with such virtual network providers. The entrepreneurial task of the big data virtual network provider is to decide over the specificities of the AI system implemented according to the requirement of the physical IoT applications. In addition to the required data privacy and security regulations the elaboration of new liability rules for AI interacting with traditional technologies is becoming relevant taking into account AI-specific ethical and transparence obligations.