Berlin, 04.11.2022
1) What is the objective of the Data Institute?
There is already plenty of data in the current ecosystem, but it is neither centrally collected and made accessible, nor is data usage cooperation between companies actively incentivized. This leads to the fact that a large part of all available data is not used or cannot be used. We are therefore convinced that the Data Institute can make a significant contribution at this point, both in terms of increasing the visibility of available data and in terms of standardization. One possible goal of the Data Institute could therefore be to make available a central data catalogue from public administration, research and the private sector.
Another possible goal of the Data Institute should be to take a pioneering role in Europe in promoting the creation of learning datasets for large AI language models (e.g. OpenGPT-X) on the one hand, but also to make them available in a bundled form. While learning datasets in English are currently the standard, we are convinced that with the initiation of targeted initiatives, German datasets can be promoted and new standards can be established at least for the domestic AI ecosystem.
2) What are the most pressing problems that the Data Institute could help solve?
The current biggest challenges in the areas of data use and data sharing can be categorized into three areas: Visibility of availability, standardization of data, and legal uncertainties.
We strongly believe that the Data Institute can bring about progress in all three categories. First, the creation of a central data catalogue can greatly increase the visibility of data and create a one-stop shop for research and startups. Such a step will increase the efficiency of startups and minimize the risk of multiple data collection.
Such a data catalogue also establishes an exciting idea with regard to a possible taxonomy, similar to an ISO certification. The Data Institute can make a further contribution with regard to the standardization of data and interfaces. As a possible third goal for the Data Institute, we see the establishment of a central advisory office for startups on how to deal with guidelines, processes, or compliance with the General Data Protection Regulation (GDPR) regarding data use, sharing, and collection. As an example of orientation, the recommendations around the IT-Grundschutz by the BSI can be used here.
3) Which target groups should the data institute address?
The Data Institute should primarily address two target groups: the public administration and the startup ecosystem.
With regard to the public administration, the Data Institute can play an intermediary role, especially in areas where public authorities are not allowed to share data with each other due to applicable guidelines, for example the Online Access Act (OZG). Not only can the Data Institute act as an intermediary between authorities, we also see the potential for processing data from the public administration for general use. Furthermore, the Data Institute can promote transparency with recommendations; we are thinking here, for example, of research funding.
A second target group of the data institute should clearly be start-ups. With the creation of a central data catalogue and a central contact and consulting point, value creation potential can be realised and competencies bundled.
4) Which tasks should it fulfil? And which explicitly not?
Central data catalog:
For startups, but also for research, access to data is of enormous importance. By creating a central collection point, processes can be significantly simplified. The Data Institute can take on the role of a broker here, a role that can be of central importance with a view to the commercialization of data.
Provision of sensitive data:
Particularly for making sensitive data available, such as from the health sector, government agencies are the obvious choice. While clinics currently have to collect and clean up data individually, the Data Institute can take on the role of a central collection point in which data is prepared for the general public.
Central Counseling Center:
Until the legal uncertainties, especially for startups regarding data use and sharing, are clarified, there is a need for advice and support in dealing with the GDPR. We are convinced that the Data Institute, for example oriented towards the services and recommendations of the BSI on the topic of IT-Grundschutz, can take on a supporting role to further strengthen the innovative power of German AI startups.
5) How can the Data Institute best accomplish these tasks?
In order to be able to fulfil the possible tasks formulated above in the best possible way, it is imperative that the Data Institute can take on a central role, set practice-relevant standards and promote innovative initiatives. We therefore recommend that the Data Institute formulate relevant use cases as a first step. Furthermore, we are convinced that the work of the Data Institute and its formulated guidelines must be applied to both lower and higher-level authorities. Only in this way can a standardization of data and interfaces take place in the area of public administration.
6) Are there any aspects that are particularly important in the organisation and structure of the Data Institute?
For the structure of the Data Institute, we appeal to the principle of “form follows function”, by which we would like to express the goal that the concrete goals of the Data Institute should be formulated in a first step, and the organizational structure should subsequently support the achievement of those goals. We expect the Data Institute not to assume the role of a regulatory authority, but rather to act as an innovation-promoting actor in the German data ecosystem and to break down existing barriers. We therefore expect that both the goals and the structure of the Data Institute will be oriented towards the demands of practice.
About the KI-Bundesverband
The Bundesverband KI links the most innovative AI and deep tech companies with established business and politics and, with more than 400 AI companies, is the largest AI network in Germany. The members of the Federal Association for Artificial Intelligence are committed to ensuring that this technology is applied in the spirit of European and democratic values and that Europe achieves digital sovereignty. To achieve this, the Federal Republic of Germany and the EU must become an attractive AI location for entrepreneurs, where willingness to take risks is appreciated and innovative spirit triffts on the best conditions.