Institutions, Data Access and Use, 5 credits

We are entering an Artificial Intelligence (AI) era. Machine Learning (ML) algorithms thrives on access to big and varied datasets. Whether produced by human or sensors, actuators, drones, robots, or other “things”, the access to datasets is crucial for innovation and operational efficiency. Industry practices related to access and use of the machine-generated data are hitherto largely ad hoc: neither a legal framework, nor business norms are yet in place. Current data access practices are heavily ruled by bilateral contracts and technical protection means and the former has limits in up-scaling while the latter introduces factual exclusivity. Technology-wise, AI data is a “moving target”, highly dynamic in volume, variety, and velocity (3Vs). This dynamic also goes for the actors – enterprises, machine-owners/manufacturers, and consumers – who create that data. That causes problems, in allocating access and control, in a networked environment, as the value of data is often hinged with the access of others’ data. It is not evident who should access and when to access particularly when data is a combination of information in varied public and proprietary domains.

This course introduces institutional, industrial, and organizational factors that are relevant for AI data access and coordination. The course participants are most likely the various levels of business decision makers, including executive/project managers who are dealing with inter/intra – organizational communications at their daily bases (and legal counselors within the firm).

To understand the AI data coordination mechanism, fundamental governing factors are to be considered. The ongoing emergence of internet-of-things takes place in an environment where rules for access to and use of machine-collected data are uncertain. Legislations and standards for data trading are under development and not yet in place and there are divergent views regarding legislation on what it should look like. Key issues include: (1) whether and how the design of intellectual property rights stimulates knowledge creation and promote trade, (2) how machine-generated data/innovations are treated in tort (liability) cases where an act or omission give rise to harm of another, (3) how competition is affected by new contractual relationships and by (platform-based) economies of scale and network effects.
In this course you will learn how data are accessed in the new economy of AI and ML and what kinds of institutional, industrial, and firm-level factors are governing AI data coordination and (re-) use. This is important, not least at the arrival of the 5th generation of wireless cellular telecommunication technology (5G), whereas an estimated 75 billion devices are connected (in 2025) and there are oceans of data (libraries) that business, engineering, and financial processes can draw on. The focus of this course is industrial data, namely business-to-business (B2B) data, but other types will also be covered.

This course offers a snapshot on the complexity of the problems involved in of AI data access and sharing, and on the factors affecting that sharing. This course provides a link between the institutional economics and industrial management approaches on AI data coordination and appropriation (i.e., with a focus on the interplay between institutional change and corporate behavior in the new economy).