A Digital Single Market for Europe: the data economy - Main contents
Purely in financial terms, by 2025 the data economy of the EU-27 is likely to provide 5.4% of its GDP, equivalent to €544 billion.
Data is the basis of everything digital - and now potentially as important to business and society as the internet itself.
While raw data has an intrinsic economic value, its full and true value comes only when it is interpreted into useful information and used in different ways. There is far more than one use: it can be marketed and re-branded, re-used, aggregated, transformed, bought and sold.
However, that huge potential is limited if data cannot move freely.
It should be able to migrate between locations, across borders and within a single data space.
In 2014, it was clear that direct and indirect barriers that EU countries imposed on using and transferring data were holding up data flows and constraining the entire EU data market.
Just by removing these restrictions, we estimated, could generate up to €8 billion in GDP a year. That was our aim with the regulation on the free flow of non-personal data: to allow more people and businesses the chance to make the most out of data and its opportunities.
This law applies from May 2019 and prohibits EU countries from putting laws in place that unjustifiably force data to be held solely inside national territory.
When we examined this market, we found 56 different laws in effect in 2014 across 21 EU countries that forced data localisation.
By 2020, there should be none.
The EU is first in the world to have this rule. It increases legal certainty and trust for businesses. It makes it easier for SMEs and startups to develop new innovative services and enter new markets.
Getting the most from public data
Then we looked at how to make more public information and data available for use and re-use. Public sector bodies produce and possess large amounts of data. Here, there is huge potential value to gain by adding innovation value to the original data gathered.
This includes traffic data, geographical and weather information, general statistics, education and social data, data from publicly funded research projects, certain library books, and so on.
But there are several barriers that make it hard to re-use public sector data for commercial purposes. Real-time access to public sector data is relatively rare, and can cost a lot for re-users: too much to make it useful for startups and other smaller users.
In Sweden, Seapilot produces navigation apps based on marine chart data. In the US, this publicly funded data is freely available for re-use.
In the EU, costs for the same datasets vary from €2,745 in France to around €18,000 in Italy.
It does not have to cost so much. It could be free.
Today’s rules do not cover enough data generated by energy and transport utilities, or publicly funded research - even though much of it is fully or partly funded by public money.
So we updated the EU’s 2013 law on public sector information to make it more available - in real time too - and reduce charges for re-use. In the future, high-value datasets will be available for free.
The updated rules should help to boost EU growth related to public sector information from €52 billion in 2018 to €194 billion by 2028. Without them, it would be a lot less: €150 billion.
Getting access to large datasets is important for a wide range of business communities.
Scientists and researchers mine and analyse it to develop new knowledge and insights, often by discovering correlations between materials produced in different scientific fields.
While text and data mining (TDM) is an emerging and important tool, it has developed quite slowly in Europe - mainly due to legal uncertainty caused by EU copyright rules, in particular because of complex rights issues and high compliance costs.
That has now changed.
The EU’s copyright reform simplifies the rules for using TDM on large datasets for research and other purposes, education and preservation of cultural heritage.
This should boost the development of data analytics and artificial intelligence (AI) in Europe.
Digitising Europe with sheer computer power
With constant rises in data volumes and flows, it was clear to us that European industry depended on having access to high-performance computing (HPC) as it turned more digital.
We were falling behind. We needed to do more to develop national HPC ecosystems, get exascale supercomputers up and running - capable of performing 1018 operations per second, a quintillion - and to support the European Open Science Cloud.
Visiting Barcelona Supercomputing Centre, February 2018
As scientific and industrial challenges grow in scale and complexity, HPC users are seeking greater computational power.
Without it, European scientists and industry will increasingly process their data outside the EU, undermining the growth prospects of our digital economy.
Today, industry in the EU consumes over 33% of supercomputing resources worldwide but only supplies 5% of them. Germany recently entered the list of the world's top ten supercomputers - in eighth place - with a Lenovo-built supercomputer near Munich.
As part of the DSM, we wanted a strategy to provide researchers, industry, SMEs and public authorities with access to integrated world-class supercomputers.
In practice, this is a scheme - a joint undertaking - to pool European resources to develop exascale supercomputers for processing big data based on competitive European technology.
The plan is to place a European supercomputer in the world’s top three by 2022. From next year, two common-owned supercomputers will be available for business and research users.
Increased data flows and quantities, stronger data protection: all these are requisites for Europe to develop and advance in future-oriented technologies such as the Internet of Things, AI, robotics and data analytics. Data is the raw material that makes them work.
Artificial Intelligence for Europe
With most AI technologies, for example, the larger a dataset, the more and better that they can learn. Our DSM work on data paves the way for Europe to catch up ground from global competitors based on ethical guidelines for AI use agreed by a group of experts.
We will be developing common data spaces in areas such as health, energy and manufacturing to aggregate data for public sector and for business-to-business.
In AI, international competition is fierce. Both China and the United States are competing to become a world leader in this area, and both are pumping large amounts of public money into AI research. Tech giants are also investing billions into AI research and innovation.
Along with all EU countries drawing up a national AI strategy - which they are doing - what is urgent now for Europe to commit to much more funding and investment in AI.
Our targets are for at least €20 billion of investment in research and innovation by the end of 2020 and more than €20 billion per year over the next decade.
Funding under the EU’s long-term budget, running from 2021 to 2027, will also help to achieve these targets and complement national investments: the EU has proposed investing at least €7 billion from the Horizon Europe and the Digital Europe Programme in AI.