Europe's future economy boosted by data exchanges
Interview of Malte Beyer-Katzenberger, Policy Officer at the European Commission
With the Data Governance Act and the Data Act, and the Gaia-X initiatives, Europe is illustrating its leadership in developing and building a strong and regulated data economy. Creating secure data exchange frameworks brings trust, traceability and strengthens data governance. By facilitating the exchange of data across geographies and industry sectors, public and private organizations gain powerful information to foster innovation, address environmental, social and economical challenges and ultimately create an impact on our society. At the 2022 CDOIQ Symposium in Lausanne, we sat down with Malte Beyer-Katzenberger, Policy Officer at the European Commission, to further explore the European Data Strategy.
Dawex: What are the main elements of the EU data strategy?
Malte Beyer-Katzenberger: The strategy from 2020 was written under the impulse of Thierry Breton, our industry commissioner and internal market commissioner. And he sees a great potential to use industrial data. As a second wave of data that have been currently under-exploited, we have seen a big exploitation of personal data with also the mega platforms that we know. But as we are now seeing the data dividend coming to industry, we also need to work on leveraging this potential of industrial data, which is mostly non-personal. And one big group, we will see this probably later is then data from connected devices, from the so-called internet of things. The strategy really wants to work on reusing existing data. Maybe also pooling data because data exists and is often underused.
80% of the data that we have today is underused, sometimes because there is no better use. But what we want to do is to really reuse existing data for purposes that maybe have currently not been thought of, or for these data pools that we need also to train and the AI in the volumes that we need. That's maybe the first big block.
The second element is that we have already done a lot of work on government data. So we've tried with the open data policy to open up data that is held by government, which is not in the big data amount that we see in industry. It's really structured data. it's often static data, but still interesting data, it can also be weather information. So weather services have large amounts of data. We have worked over 20 years now to really make this data reusable. That will be again, another step with these so-called high value data sets, which means that really the restructuring data are available at zero cost in a good quality format. We're working on this at the moment.
But the challenge, I think today, is to really bring data collaborations between industries, so that the business data that exist are also exchangeable, reusable, poolable if you want.
The data strategy here says first and foremost, it should be driven by market mechanisms. And only as a last resort, in sectors where there is a competition problem, you may go for an obligation on some of the operators to open up the data. So that's the second thing, really trying to have a market-driven approach because it needs to make business sense to exchange the data, but also then maybe to nudge companies, and we'll see this: to exchange more data.
The third element is that we have to think Europe, we have to think large, we have to have a data space. The general data protection regulation is not only about protecting data. It's also about creating the free flow within the Union. That's often very much forgotten. So that we have a level playing field for personal data.
What we intend to do is also to make sure that non-personal data can be reused across the continent so that we have that same space that is sometimes the “maybe advantage” of a competitive economy across the Atlantic. And then bringing also industrial collaborations that are no more national, anyway. And that's really the big second. So, a single space for data, that's the third element. And the fourth element is more like, how are we going to achieve this? And one thing is to lay down legislation, that's a classical role of the European Commission. But the other important role is also to use funding programs, not only the research funding program, Horizon 2020, but also a new digital deployment program, which we call the Digital program, to fund activities that deploy mature digital technologies. And one of the features is to fund what we call the Common European Data Spaces, which is a big, big, I think, building brick also of our data strategy.
Dawex: What are the specific role of data spaces and the role of data intermediaries in this context?
M. B.-K.: So for us, the data spaces are a structured way of organizing a data ecosystem. I mean, data ecosystems emerge naturally, whether it's a use case. What we want to do is to stabilize this more ad hoc evolution of data ecosystems with the strategic objective to really lower the cost of getting ecosystems established, have a rule book at the legal, but sometimes more contractual legal and at the technical level to say to businesses, this is the way you should collaborate, because this is basically bringing the balance between interests of creating use cases, so sharing, but also the interest of protecting the value that is in new data and protecting maybe also the information.
So the data spaces at which we're currently working on in different sectors shall bring the mixture of legal and technology to structure ecosystems and have a template to bring new use cases more quickly than we see it today. Because today we see a lot of experimentation, which is very useful. I should mention maybe Gaia-X, and also the reference architecture of the International Data Spaces Association, which is industry-driven, coming mostly from the manufacturing industry and maybe adapted for industrial data collaborations. But maybe we need to adapt them also for other situations, let's say tourism or finance, where other problems occur in terms of secrecy or cultural heritage elements being in there so that we're really capturing all possible issues that can happen around data and that we bring then ecosystem together.
We have a template to bring it. But it has to be use case driven. And I want also to maybe rebut one, maybe very common misconception, a data space is not about a common data pool. It's certainly not a common open data pool. It's more a framework within which industrial business data exchanges can happen if the businesses want it so. If they find a partner to work on data, if they find companies to pool the data with, it has to be ultimately use case driven and has to be voluntary. But with the kind of the toolbox to make these collaborations happen, we think it's going faster. And intermediaries or data exchange marketplaces will play a very critical role for us because they will professionalize the situation of exchange, which you can also work on ad hoc, and you can establish.
These companies have done so for many, many years, and they're doing it. But with the framework of the data spaces, plus the service of exchange and the technology of facility in exchange, we believe it will go even faster. That's why these intermediaries, we've tried to capture it in the Data Governance Act, which was the first legislative piece under the data strategy to say we need a framework for a different type of intermediary. The crucial difference between a data broker, as we knew it, and the Data Governance Act intermediaries is that the Data Governance Act intermediaries are really at the service of those who want to share data, but where the data holder of which wants to share the data wants to capture the value himself, and also keep a certain level of control. This means value capture is very important. This means that we change, well, we impose, so to say, a very kind of specific funding model for the intermediaries, which is not to speculate on the potential value of the data that they exchange and say, I can play with this data myself and I can monetize them independently of the data holder. But I'm rather only there at the service of the data holder and I will be paid a commission transaction fee or a subscription fee, but I will not derive myself value from the data. All the value from the data should stay with those who want to exchange because this is one of the elements where we think people hold onto their data because they fear to lose value because they can't really know how much it is in value. And they fear if they hand it over to a data broker, all the value would be with those agents. They would get rich, as also some platforms in the personal data economy got very rich because they could better ascertain the value of the data. We're not saying this other model shouldn't exist anymore, and in the non-personal data economy it can also continue to exist, but we want to put next to this model of the traditional data platforms, the model of the data exchanges or the data intermediates as a service provider, which has a model based on a monetary compensation for the service they provide, where all the value from the data that will ultimately result stays with the data holder.
Dawex: What is the role of the Data Act?
M. B.-K.: We worked in stages for two reasons, as a reason of substance and the reason also of political process. The Data Governance Act is really about the intermediary structures. We've spoken already about the data exchanges, the data marketplaces, the facilitators maybe of data spaces as the kind of examples of data intermediaries. But on top of that, we also work on intermediaries for public sector information. We, if you know, the French Health Data Hub, as one example of facilitating research on existing public and private health data. Fin Doctor is another inspirational example from Finland.
Then also we work on altruism intermediaries. So we have actually three types of intermediaries in the Data Governance Act. So the Data Governance practices really about what some call data institutions. If you look at the work of the Open Data Institute from the UK, there's a lot of work on what they call data institutions. You may have heard the term of data trust. That's certainly where a specific type of intermediary where you entrust a third party to monetize data on your behalf, but in a trust relationship, in a not only trust in a sense of I trust you, but also in a sense of a financial trust situation. So exploit my data on my behalf, but give me the most share of it. That's certainly also a subset of data intermediaries that we really see, and that came out of the French and the UK AI strategies, which really inspired us to do.
Now the Data Act, we cut off all questions of substantive rights on data, which is really what chapter two of the Data Act is about. The Data Act has more things on cloud, switchability, has things on fans of contract and even on the use of data by government. So, there's four to five big policy topics that are also emerging new policy topics, or are pressing competition concerns when it comes to cloud. So, we cut them a bit off to make it hand manageable in the political processing in the council of ministers and in the parliament.
But chapter two of the Data Act facilitates data exchanges or will make data exchanges happen more often because more operators will have a right to receive data. What we see in some parts of the IoT data economy is that manufacturers of the objects use the technical defect to control they have over the data, because they normally construct the architectures in a way that the data land on their service first, that's very prominent in cars and in farming. To also then basically have the first possibility to offer services on top of this or team up with service providers. That gives them a very comfortable position to digitize. It is also a very useful position to offer new services. But we feel that the operators of a connected device, that can be a farmer who runs a tractor, can be a construction company who runs a construction engine, can be a manufacturer who has a factory of robot or similar or even a consumer that they always should receive a copy of the data that they create when they use the object.
So then we have two kinds of parties who have a right to the data, which will multiply, in our view, the types of services. And that is then, again, something where you can go to a data exchange or a data marketplace, or a data space to say, now that I have a copy of the data that my engine generates, I can also go independently of the manufacturer to find myself a service provider. I think the possibilities to use and monetize non-personal data multiply because more players have a right to actually be active in the data economy. And for us, this is an important stimulus for competition and innovation on these. I wouldn't call them aftermarket, it's adjacent markets to connected objects, which is really also, I think, a coming frontier in industry and also on the household side. We use a market-driven mechanism, which is the portability, right? I'm operating a device, I have the right to receive my data, import them to a service of myself.
It's not that we, the European Commission or the European Union, decide which operators should have access to which data. But it's only if the user, the operator of the object wants it so, he has an economic right to actually be active as an economic agent on the market for disconnected IoT data.
And that's really what the Data Act brings on top. And we've tried to distinguish it from the Data Governance Act so that we have institutions and processes in one act and substantive data rights in the other.
So thank you very much for the opportunity to explain the EU policy. I want to say this is a policy that we do together with the member states. So it's really also collaborative at the European scale. But we're driving this in order to stimulate a European data economy.
Explore the benefits of data regulations with Dawex experts
Want to hear more about the benefits of Data Exchange?
Subscribe to our newsletter here.