Statement of the KI Bundesverband on the revision of the Data Strategy 2021

Berlin, 04.11.2022

1) What are your expectations for (further) development?

We welcome the Federal Government’s implementation of the data strategy to date. Even though progress is being registered in important areas, it must nevertheless be noted that implementation is progressing more slowly than expected. The establishment of data laboratories in the federal ministries, for example, is undeniably a step in the right direction, but this measure ultimately forms only a subset of an adequate data strategy.

Our expectations for a new data strategy relate to three main points.

First, we are convinced that a new data strategy can profit from a prioritization of the measures already formulated and a resulting acceleration of implementation. New, innovative data use concepts, such as data marketplaces or a central data catalogue for public data, should also be considered in this context. We also call for coordination of the data strategy with the Data Institute, which is currently being set up in parallel and is not being thought through together enough.

Secondly, we expect a commercialisation of data to be discussed and concrete measures to promote commercialisation to be presented. There is enormous value creation potential in data for Germany, especially for the German startup ecosystem.

Thirdly, start-ups in particular face legal uncertainties regarding the use and sharing of data. We therefore expect the new strategy to reduce legal uncertainty and provide a central point of contact to advise companies or research institutions.

II. How can we better leverage the potential of data (“new oil”)? In which areas do you see the greatest potential?

2) Where do you see hurdles in data access/use? How can these be reduced?

For innovative startups, access to data is of enormous importance. Only if these are available in good quality, i.e. without serious distortions and sources of error, and in sufficient quantity, can qualitative AI applications and products be developed. However, the availability of data often fails due to a lack of access to the data pools of large international corporations and increasing monopolization tendencies. In areas where data is available, however, the applicability for AI often fails instead because the data are not compatible with each other and a central broker is lacking.

In addition, there is a lack of a central catalogue of publicly available data that would enable new innovations, especially in the public sector, and save companies from having to collect data independently and at great expense. Often, due to the lack of opportunities for data sharing in research and business, the same data sets are collected more than once. Here, too, the availability of public data sets would help to avoid inefficiencies in data collection. In our view, the Data Institute would be a suitable partner for the compilation of such a catalogue.

Last but not least, there are still legal uncertainties for startups with regard to data use and access within the framework of the General Data Protection Regulation (DSGVO). However, legal hurdles do not only exist with regard to the DSGVO, barriers are also produced by the Online Access Act (OZG). We therefore expect that the regulations concerned (e.g. DSGVO or OZG) will be readjusted.

3) What tools or measures would make it easier for you to access data?

As mentioned in the previous question, the guarantee of data availability currently fails, for example, due to immense data monopolies, incompatibility of data, or simply the lack of a central marketplace. This problem is exacerbated by the unwillingness of established companies to enter into cooperative data usage models. We therefore recommend the creation of targeted incentives to make cooperation between large companies and start-ups more attractive and to counteract those data monopolies.

We are convinced that the awarding of government bonuses to incentivize such cooperation models is a good starting point for facilitating data access. The AI voucher model, for example, presents itself as a possible concept.

The establishment of a data marketplace to create commercial incentives for the monetization of data could also simplify access to data. Here, coordination with the Data Institute, which could possibly take on such a task, would be desirable.

We also recommend the promotion of open data pools for the benefit of education, research and industry. In these secure, uniform and safe open-data pools, data from both public administration and industry can be made available in an uncomplicated manner.

4) Where do you see concrete potential for regulatory improvement?

Addressing “commercialization of data use and sharing”:

In the area of commercial use of data, or the monetization of data, there is currently a certain regulatory leeway, which is why we consider the establishment of a data marketplace to be welcome. Coordination with the Data Institute is absolutely necessary here.

Update on the General Data Protection Regulation (GDPR):

There are still numerous ambiguities regarding legal certainty with regard to the General Data Protection Regulation (GDPR) for companies as well as in the regulation on data access and its use. In connection with the revision of the data strategy, we therefore also recommend an examination of the GDPR and a related improvement. We consider, for example, the launch of a long-term governmental advice centre for start-ups to which they can turn in questions relating to the GDPR to be welcome. We are convinced that this regulatory improvement will increase the potential for innovation and reduce legal uncertainties.

5) Where do you see concrete potential for improvement in standardization?

With regard to standardisation, we are clearly convinced that there is not only potential for improvement here, but that there is also a clear need to catch up. In particular, collected data from public agencies are often not compatible with each other and therefore cannot be used for combined purposes. For this reason, we are convinced that guidelines for data collection and sharing must be set in a new data strategy, which can then be further specified and developed by a data institute, for example.

6) Where do you see concrete potential for improvement in interoperability?

For the public administration in particular, there is great potential for improvement in the area of interoperability, which could be addressed with a new data strategy. In our opinion, the Online Access Act (OZG) creates clear obstacles here, which at least complicate data exchange between authorities, if not make it impossible. Such a hindrance only promotes the inefficiency of state authorities, which is particularly problematic in the current times of a general shortage of staff and specialists.

By amending relevant ordinances, as in the case of the OZG, public administration can be continuously modernised and digitalised and administrative efficiency promoted.

III What do we have to do for this?

7) What new ideas/ideas from other EU countries would you like to implement with data and what is stopping you so far/what would have to change for you to be able to tackle this?

Based on a workshop with association members, we would like to bring in a specific use cases and highlight necessary changes in a new data strategy.

Example: Sharing and use of sensitive data

If start-ups or research institutions want to carry out studies for which they need data from the healthcare sector, it is necessary to discuss the data with all the clinics involved.

On the one hand, this complicates the work of conducting studies immensely; on the other hand, data are also frequently available only to varying degrees (individual ethics committees, differenced data collection, etc.). In the UK, this work is undertaken by the National Health Services (NHS). This allows data to be collected centrally and standardized.

The example of the NHS is exemplary for a number of such use cases. We are firmly convinced that the potential for value creation can only be realized if data is standardized and made available centrally. Government agencies are an obvious choice for making sensitive data available, such as that from the healthcare sector. Here, too, cooperation with the Data Institute would be desirable.

8) What priorities should the 2022 data strategy include?

The priorities of a new data strategy 2022 are clear to us:

Firstly, the measures already formulated in the 2021 data strategy urgently need to be fleshed out and prioritised.

Secondly, we appeal for responsibilities to be clarified and synergies to be pooled. The separate conception of a data institute and the revision of the data strategy lead to a double discussion and ignore the numerous overlaps. We therefore call on the federal ministries to cooperate more efficiently in such processes. For this reason, the tasks of a data institute should be included in a new data strategy.

Thirdly, a catalogue of measures to clear up legal uncertainties in connection with the use and sharing of data is recommended, especially for start-ups. Current obstacles and uncertainties with regard to the GDPR must be reduced and a central advisory office for start-ups must be created.

Fourth, we expect data commercialization to be addressed. If a new data strategy can lay the foundation for a data marketplace, we are convinced that such a marketplace offers significant value creation potential for Germany.