A bimonthly magazine on international affairs, edited in Germany's capital

A Question of Sovereignty


Angela Merkel is deeply worried that Germany and the EU will fall far behind the US and China behind in developing AI. So far, however, her initiatives have failed to produce any significant results.

© REUTERS/Kevin Lamarque

Whenever Angela Merkel considers a political issue to be of pressing importance, she invites experts to the Federal Chancellery. The evening of May 29, 2018 was no exception: Merkel convened a group of 20 experts from business and science to discuss artificial intelligence. The chancellor has long been worried that Germany and the EU may miss out on this groundbreaking technology.

To also give her cabinet the benefit of the specialists’ urgent message, Merkel asked a number of ministers to attend, including her close confidant, economics minister Peter Altmaier, and labor minister Hubertus Heil of her coalition partner, the SPD. Germany is meant to have an AI strategy by the end of the year, so the cornerstones of a strategy have to be in place before the Bundestag’s summer break in July.

The chancellor has long been clear: Dominance in AI could entirely restructure the geopolitical order—or serve to cement American supremacy and Chinese power. There is resistance to this latter outcome, above all in Berlin and Paris. If there is one thing that French President Emmanuel Macron and Merkel are in absolute agreement about, it is the strategic importance of AI.

Since taking office, Macron has been talking of Europe’s need to defend its “strategic sovereignty”, adding that this will be a battle fought on many fronts, including technology. In a June 3 interview with the conservative Frankfurter Allgemeine Sonntagszeitung, Merkel emphasized that as early as 2017, she has been calling for Europe to take its destiny more into its own hands. Most have taken those statements as references to Europe’s security dependence on the United States. But she was also taking about technology.

It was in 2013, in the wake of the NSA spying scandal, Merkel first had to acknowledge how being dependent on American software and Chinese hardware could limit a country’s capacity to act. Five years later, during a visit to China, Merkel took a detour to the high-tech city Shenzhen, where she visited the Chinese startup iCarbonX and saw how far advanced China is in the usage of big data for the health sector. It was a painful reminder for a chancellor who, for all her time in office, has yet to witness a successful introduction of an electronic health card in Germany.

As always after a trip to China, Merkel returned to Germany impressed and worried by the pace of change. While she is quick to stress that certain aspects of communist China—be it the desire to control citizens or the lack of data protection—could never be replicated in in a democracy, that doesn’t change the fact that German companies have to compete on the same field as American and Chinese tech giants.

Ever Louder Warnings

Warnings about the explosive potential of AI development have been getting louder and louder. Military officials believe that AI is already in the process of revolutionizing warfare. The US and China have been long been experimenting with autonomous weapon systems, drones, and ship fleets that can organize themselves. While Germany was debating whether its conventional weapons systems are even operational, or if there are enough boots and protective vests for its soldiers, other places were developing the technology that will decide future wars. The technological gap is widening even more rapidly—and in the outdated German security debate, nobody seems to have noticed.

Merkel on numerous occasions, even to her coalition partners, has warned of the need for increased spending on German defense. But she has so far avoided a debate about the possibilities and dangers of these new types of weapons. In the meantime, Germany’s Foreign Office has begun to consider how AI will change foreign policy. An international debate over the ethics of using automated and autonomous weapons has begun.

Merkel has, to this point, limited her attention to the shortcomings of digitalization in the civil sector. She has been pushing the issue since 2012. Her initial efforts attracted ridicule after she described the unpredictable consequences of the internet for society as Neuland—terra incognita. Then she threw her weight behind Industry 4.0, a campaign meant to push German companies to understand and implement the digitalization of production, administration, research, and sales. Merkel warned that, unless Germans master Big Data and the specialized manufacturing technology of digital-age products, proud German industry could simply become the work bench for American IT firms.

In 2016, the term “artificial intelligence” finally appeared in the chancellor’s vocabulary when Merkel, a trained physicist, realized during conversations with entrepreneurs and scientists that artificial intelligence would massively accelerate the fusion of otherwise separate research domains. “I’ll put it bluntly: It is not entirely clear to me in which fields we are actually top-notch; in which fields we need to buy more knowledge; what the interconnection of diverse fields will look like one day; and if we will, by then, have all of the technological capabilities that we need to be at the front of the pack,” she acknowledged at a research summit on June 12, 2016. Afterwards, Merkel called even more urgently for Germany to go on the offensive in AI research. If necessary, Germany should protect its AI companies from being taken over by American or Chinese firms.

Three Competing Ministries

One reason Merkel was turning up the heat was her improved understanding of the complex effects of AI, which she now describes as “disruptive” or “revolutionary.” Another was the realization, at the end of the 2017 legislative period, that the grand coalition’s digitalization efforts have had little success. For example, though national broadband expansion was ranked among one of the government’s top priorities in 2013, Germany’s position in the international rankings for data connectivity has continued to worsen.

The chancellor credits herself for the steady expansion of Germany’s research budget. But research expenditure was not tax deductible—an issue especially important small and medium-sized businesses—until 2017. Venture capital has also been slow to arrive. In the last election campaign, politicians of all stripes complained that having three ministries, run by three different parties, responsible for digitalization didn’t exactly speed things up. On top of that, critics said that the government had made too many special considerations for Deutsche Telekom. By allowing Telekom to increase network speeds by “vectoring” preexisting copper cables, they cut costs in the short run, but only delayed the expensive and inevitable transition to fiber-optic cables.

In response, Merkel insisted during the campaign that the chancellery should centrally manage all digitalization activities in the future. In her new grand coalition, she has two senior figures dealing with the issue: Helge Braun, head of the federal chancellery, and Dorothee Bär, federal government commissioner for digitalization. Additionally, she created a new department responsible for digital issues in the central government office and integrated it with other departments from the interior ministry. In short, Merkel wants to speed things up. Her May 2018 meeting with AI experts only reinforced the point that Germany was lagging behind and needed to take decisive action. One element is to make it clear to small and medium-sized companies that if they want to survive, they have to engage with AI technologies.

In the battle to win back some technological sovereignty, lots of levers need to be pulled at the same time. That is not easy given Germany’s federal structures and the distance between politics and the economy. Unlike Beijing, Berlin cannot “rule from the top.” And unlike in the US, there are not huge sums of private money ready to be spent on scaling up start-ups or accelerating IT research. In Germany, the federal government has to ask the states to reform their educational systems in order to meet new technological challenges. And when the big companies in China or the US come calling, company managers naturally think about their own interests or the firm’s interests—not necessarily about Europe’s strategic needs.

The Complete Value Chain

In the tech age, a region only has technological sovereignty if it can produce the complete value chain of digital products. Whether it be computer chips, computers, batteries, or software, the European countries have collectively given up on competing at the top. Instead, they have become customers for other nations’ companies. The US and China dominate the market for software, hardware, and social media platforms, which have an ever-greater impact on daily life. And when there are interesting technological developments from German startups or companies, the large American and Chinese firms eagerly buy them out.

These problems are why the 2016 Chinese acquisition of the German robotics company Kuka generated such a passionate debate in the government about foreign investors. Should foreign takeovers of strategically important companies be more strongly controlled? “We need to exercise caution, so we can maintain a foundation, so that not everything will be bought out from us,” warned Merkel in May 2017. Merkel also called for more flexible European aid rules that would allow member-states to better support AI firms.

Yet the attempts to redress Europe’s deficiencies seem modest given the speed of innovation elsewhere. For years, Merkel has worked in the background with some similarly minded European leaders to try and develop an independent European computer chip factory, or perhaps a European battery factory. But the fragmentation of the EU internal market, national reservations, and the lack of strategic direction have hindered progress. Only in 2018 did the European Union implement its General Data Protection Regulation, which provides a common legal framework for handling data in the EU. Everything is too slow.

A Three Percent Benchmark

Merkel and Macron want to kindle a new research dynamic—or at least expand on current basic research and its applications. For the chancellor and the French president, the possibilities of AI are so revolutionary that the research needs to be revolutionary too. The model is the Pentagon’s DARPA program for defense research which has regularly boosted civilian R&D. But accepting that research projects will sometimes fail conflicts with the German approach of accounting for every penny spent.

The leaders of France and Germany argue that such a cautious, rigid approach makes it impossible to discover the necessary “game changers” or “technological leaps” that could secure the survival of European industry. They believe that the EU should use the European Innovation Council to support high-risk research projects—even if nine out of every ten projects will ultimately fail.

It is not clear whether the drive to catch up will work. China and the American tech giants have been following forward-looking strategies for years, investing tens of billions in new technologies. In an aging Europe, on the other hand, the debate focuses almost exclusively on the distribution of social benefits and fears of migration. The EU will still have around 450 million citizens after Brexit, but no functional European digital market to serve them.

On top of that, there is the general lack of awareness about the importance of innovation. Much of the EU is absorbed by passionate debates about NATO members’ commitment to spend two percent of GDP on defense. Yet nobody seems to be noticed that almost all EU member states are seriously neglecting another, more important commitment that dates back to 2010: they should all be spending three percent of their economic performance on research and innovation.