Use of the technology is picking up pace
Photograph: Siemens
Listen to this story
Your browser does not support the <audio> element.
A cross the world, most businesses are still merely dipping their toes in artificial intelligence. Not so Schneider Electric, claims Peter Weckesser, who oversees the digital efforts of the French maker of industrial equipment. It has around 100 applications of the technology already in operation. Morgan Stanley, an investment bank, reckons that these will result in around €400m ($470m) in annual savings for the company this year. That amounts to less than 1.5% of its total costs. But Mr Weckesser has his sights set much higher. Eventually, he declares, “there will be not a single product or function at Schneider Electric that will not be affected by artificial intelligence.”
From cloud-computing to 5 G, European companies have tended to be slow in adopting new technologies. Many feared that the same would happen with the most recent wave of so-called generative AI. Yet, as Mr Weckesser’s enthusiasm shows, there are reasons for optimism.
The opportunity is big. Europe has a large industrial base and it is hunting for ways to boost economic growth. It may have fallen behind in the race to produce cutting-edge AI models—it built just three of them in 2024, compared with China’s 15 and America’s 40. And it has no hyperscalers of its own to pour vast sums into building data centres. But when it comes to widespread productivity gains, adopting AI will matter more than selling it.
Chart: The Economist
Already Europeans as individuals are among the most enthusiastic adopters of generative AI. According to research from Microsoft, a software giant, 32% of them use the technology, based on a population-weighted average of European countries, compared with 28% of Americans and 16% of Chinese. Many Europeans are hopeful about its promise. Pew, a pollster, asked people whether they were more excited or concerned about AI. Americans topped the list of worriers; Europeans were more upbeat (though not as positive as the Chinese, according to other surveys).
A growing number of European companies are likewise now embracing AI. A recent survey by the European Investment Bank (EIB), a multilateral financial institution, showed that around 37% of EU firms use the generative sort, compared with 36% of American ones. That figure hides plenty of variation. Nordic companies have been the most eager: Finland topped the EIB ’s list, with 66% of businesses using the technology; Denmark, at 58%, came next. In Italy and Greece, by contrast, only 20% and 19% of companies, respectively, were adopting it (see chart 1). Moreover, European businesses have tended to use AI for a narrower range of functions than their American counterparts (see chart 2). Only 55% of EU firms surveyed by the EIB that were using AI were doing so in at least two areas of their business, compared with 81% of American ones.
Chart: The Economist
Europe’s manufacturers, though, are well ahead of their peers across the Atlantic. The EIB ’s study found that 48% of them use AI, including older types of machine learning and “big data”, whereas only 28% of American manufacturers do. Siemens, a German industrial giant (whose chief executive, Roland Busch, is pictured), has been using ai at its futuristic factory in Erlangen, in Bavaria, for over five years. More than 100 algorithms are improving production at the site. Schneider was also an early adopter. It hired its first “chief AI officer” in November 2021, a year before the launch of Chat GPT. Other manufacturers are eager to catch up. Last month 25 teams from Trumpf, a German maker of industrial tools such as laser cutters, came together for a “hackathon” to develop applications for the technology.
Many other European businesses are AI -curious. Carlsberg, a Danish brewer, has used it to create a tool for sales staff that, among other things, helps them choose the right promotions for customers. AI helps the firmdesign packaging and plan the transport of beer crates. Employees are encouraged to experiment with Copilot, Microsoft’s AI assistant. One new staff member who had a patchy onboarding experience created an AI tool to guide newcomers through their first month or two at the company. Carlsberg is now planning to roll it out across the world.
Some European companies are working closely with the continent’s few model-makers. In November the EU AI Champions Initiative, a lobby group, unveiled 18 such partnerships at a Franco-German summit on digital sovereignty in Berlin. These included a deal between Mistral, a French AI developer, and Helsing, a German drone-maker, to adapt AI models for defence and security, and one between Black Forest Labs, a German AI company, and Mercedes-Benz, a German carmaker, to develop tools for marketing.
ChatGPT, simplify this
What could slow the Europeans down? One risk, as usual, is regulation. Policymakers in Brussels have so far focused most of their energy on preventing the technology from wreaking havoc, resulting in the EU ’s AI Act. It sets a global benchmark for safety and data protection, but critics say it creates too much red tape and will slow down innovation and adoption. Although a full reversal is unlikely, in November the EU pushed back the implementation of some burdensome parts of the legislation after complaints from businesses. Bosses are hoping for a simplified approach that would provide clarity over what they can and cannot do. That, in turn, would encourage investment.
The other risk is economic. Already the continent’s growth is sluggish. Chinese competitors in industries from carmaking to machine tools are eating into the sales of European businesses. American protectionism is only adding to their troubles. Companies struggling for growth may be tempted to cut back their investments in AI to maintain profits in the short term. To stay competitive in the long term, though, they must do the opposite. ■
논증 분석
유형: causal
핵심 주장
유럽은 AI 모델 개발 경쟁에서는 뒤처졌지만, AI 도입·활용 경쟁에서는 충분히 승산이 있으며 이것이 생산성 향상에 더 중요하다.
논리구조
- 전제: AI를 판매(모델 개발)하는 것보다 광범위하게 도입·활용하는 것이 생산성 향상에 더 중요하다. 유럽은 2024년 AI 모델을 단 3개 개발했고(United States 40개, China 15개), 자체 하이퍼스케일러도 없지만, 도입 경쟁에서는 다른 조건을 갖고 있다.
- 논거: 개인 차원에서 유럽인은 생성형 AI 활용률이 32%로 미국인(28%), 중국인(16%)보다 높으며, AI에 대한 태도도 미국인보다 낙관적이다. (Microsoft 조사 및 Pew 여론조사 기준)
- 논거: 기업 차원에서도 European Investment Bank(EIB) 조사에 따르면 EU 기업의 37%가 생성형 AI를 사용 중으로 미국(36%)과 대등하며, [Finland], [Denmark] 등 북유럽 국가는 특히 적극적이다.
- 논거: 제조업 분야에서는 유럽이 미국을 앞선다. EU 제조기업의 48%가 AI(머신러닝·빅데이터 포함)를 활용하는 반면 미국 제조기업은 28%에 불과하다. Siemens, Schneider Electric, Trumpf 등이 대표 사례다.
- 논거: Carlsberg, Mercedes-Benz, Helsing 등 다양한 산업군의 유럽 기업들이 AI를 업무 전반에 적용하고 있으며, Mistral, Black Forest Labs 등 유럽 자체 모델 개발사와의 협력도 확대되고 있다.
- 반론: 유럽 기업들은 AI를 활용하는 기능 범위가 미국보다 좁다. 2개 이상 영역에서 AI를 사용하는 EU 기업 비율은 55%로, 미국(81%)에 크게 뒤처진다. 또한 [Italy], [Greece] 등 남유럽 국가의 도입률은 저조하다.
- 진단: 규제 리스크: EU AI Act는 안전·데이터 보호 측면에서 글로벌 기준을 세웠지만, 과도한 규제로 혁신과 도입을 저해한다는 비판을 받고 있다. EU는 일부 조항의 시행을 연기했으나 기업들은 보다 명확하고 간소화된 규제를 요구한다.
- 진단: 경제적 리스크: 유럽의 저성장, China의 경쟁 압박, 미국의 보호무역주의로 어려움을 겪는 기업들이 단기 이익 유지를 위해 AI 투자를 축소할 유혹에 빠질 수 있다.
- 처방: 장기 경쟁력을 유지하려면 경제적 어려움 속에서도 AI 투자를 오히려 확대해야 하며, 규제 당국은 명확성과 간소화를 통해 기업의 투자를 장려해야 한다.
결론
유럽은 AI 모델 개발 경쟁에서는 뒤처졌지만 AI 도입·활용에서는 경쟁력을 갖출 수 있으며, 규제 간소화와 지속적인 투자 의지가 그 성패를 좌우할 것이다.
To track the trends shaping commerce, industry and technology, sign up to “ The Bottom Line ”, our weekly subscriber-only newsletter on global business.