A Storied Past
A closer examination reveals that Europe’s contributions to AI are deeply rooted and span centuries. From ancient philosophers to modern computer scientists, European thinkers have laid crucial foundations for the field. Aristotle’s syllogistic logic, outlined in his “Organon,” is considered a pioneering exploration of mechanical reasoning. Later, Ramon Llull’s “Ars Magna” aimed to create a universal language and knowledge system, representing an early attempt at building a comprehensive AI framework.
In the modern era, European scientists and researchers were at the forefront of AI development. Alan Turing, a British mathematician, conceptualized many of the core ideas underlying modern AI. His Turing Test remains a benchmark for evaluating a machine’s ability to exhibit intelligent behavior indistinguishable from that of a human. Furthermore, the early research in AI was predominantly conducted in Europe. In 1964, the UK established the Society for the Study of Artificial Intelligence and Simulation of Behaviour (AISB), possibly the oldest AI society globally. Edinburgh hosted AI symposia for six consecutive years, solidifying Europe’s early leadership. The European Conference on Artificial Intelligence (ECAI), first held in 1988, marked a pivotal moment by separating AI as a distinct discipline from computer science. DeepMind, a European company, developed AlphaGo, which defeated world champion Lee Sedol and marked a milestone for AI. Notably, Google acquired DeepMind in 2014. This highlights both Europe’s innovative capabilities and its struggle to retain its AI champions. The early focus and significant contributions clearly demonstrate that Europe has a strong historical foundation in the field.
Regulatory Myths
Despite its pioneering history, Europe’s current AI landscape paints a different picture. A common explanation for Europe’s lagging AI development is overly stringent regulations. The sentiment “America innovates, China copies, and Europe regulates” has been circulating in various media outlets, suggesting that Europe’s regulatory environment stifles innovation. Some critics even joke that Europe’s role in the AI revolution is limited to holding meetings while the US creates and China manufactures. This perception has gained traction, influencing public opinion and even impacting investment decisions. The idea that European regulations are hindering AI progress has become a convenient, albeit simplistic, explanation for the continent’s relative underperformance.
However, a closer look reveals that European AI regulations are not as restrictive as commonly perceived. The EU’s Artificial Intelligence Act, finalized after three years of debate, is often portrayed as the final nail in the coffin for European AI. In reality, the AI Act is primarily a framework for governing the use of AI rather than restricting its development. The Act categorizes AI technologies into four risk levels: unacceptable, high, medium, and low. The higher the risk posed by an AI application, the more stringent the scrutiny and compliance requirements. Violators could face fines of up to 7% of their global revenue. This tiered approach aims to balance innovation with ethical considerations and public safety. While the fines can be substantial, they are intended to deter harmful AI applications rather than stifle innovation across the board. Blaming regulation for Europe’s AI struggles is an oversimplification, ignoring other critical factors that contribute to the continent’s challenges. The focus on responsible AI development, while important, shouldn’t overshadow the need for fostering a more supportive ecosystem for AI startups and research.
The Ghosts of the Internet Age
Europe’s challenges in the AI era are more deeply rooted in its historical experiences, particularly in the internet age. Since the dawn of the internet, European companies have struggled to compete with their American counterparts. European startups, after showing initial promise, often find themselves acquired by US firms, effectively transferring valuable technology and talent across the Atlantic. This phenomenon, often referred to as the “brain drain,” has significantly weakened Europe’s tech sector over the years. The inability to nurture and retain homegrown talent and innovation has created a cycle of dependence on foreign technology and investment.
DeepMind’s acquisition by Google is a prime example. Datakalab, a French company specializing in algorithm compression and embedded AI, was acquired by Apple. Brighter AI, which focused on anonymizing personal data in images and videos, was also acquired by an American company. Even Mistral, touted by President Macron as Europe’s answer to OpenAI, has significant American involvement. US venture capital funds and industry giants heavily funded Mistral’s initial funding rounds. It also relies on Microsoft’s Azure cloud services and has an agreement with Amazon to be a foundation model developer for Amazon Bedrock. This illustrates a persistent pattern: European AI companies, even those with groundbreaking potential, often end up being absorbed by larger American corporations. The dependence on US infrastructure and funding further reinforces the power imbalance in the global AI landscape.
French internet entrepreneur Xavier Niel warned that while Europe can currently develop promising AI models, it is uncertain whether these talents and companies will be poached in the coming years. This raises the question: what are European investors doing while European talent is being bought out? Why aren’t they supporting their own startups? This lack of local investment creates a vicious cycle, making it difficult for European AI companies to scale and compete effectively. The need for a more robust and supportive investment environment within Europe is paramount to reversing this trend.
Investment Gap
This situation highlights a historical problem that has plagued Europe since the internet boom. According to an OECD report released in May 2024, the United States leads in private investment in AI-related fields, with approximately $300 billion. China ranks second with about $91 billion, while the EU lags far behind with less than half of China’s investment, at $45 billion. European investors appear to prefer established successes over early-stage ventures. This risk-averse approach is hindering the growth of nascent AI companies that require substantial upfront investment.
In the US and China, a common startup trajectory involves a team developing a demo, securing initial funding, and aggressively expanding to capture the market, often while operating at a loss. This model, proven successful over the past two decades, is considered a necessary phase for market dominance. Venture capitalists in these regions are willing to tolerate short-term losses in exchange for long-term growth potential. However, European investors often demand immediate profitability, steady stock price growth, and dividends, even from tech startups. This forces companies to prioritize profitability over rapid growth, limiting their ability to compete with their more aggressively funded counterparts. European startups typically take two to three years to secure their first investment, while similar startups in China may fail if they don’t receive funding within a year. The faster funding cycles in the US and China enable startups to iterate more quickly, attract top talent, and gain a competitive edge.
This difference in investment philosophy impacts the enthusiasm for entrepreneurship, especially in emerging sectors like AI. A lack of funding forces companies to cut costs, leading to a shortage of AI talent and further hindering the rapid development of AI in Europe. The scarcity of capital makes it difficult for European AI startups to attract and retain skilled engineers, researchers, and data scientists. This, in turn, weakens the overall AI ecosystem and makes it harder for Europe to compete on a global scale.
The Talent Drain
The AI talent shortage in Europe isn’t necessarily due to a lack of aptitude, but rather the lingering effects of the information technology revolution, where Europe was left behind by the US and China. Many AI engineers are essentially transformed internet software engineers. The gap in compensation between Europe and the US is widening. According to Builtin, the average salary for AI engineers in the US exceeds $170,000, with total compensation reaching over $210,000 with incentives. Jobicy’s data shows that the average annual salary for AI engineers in the UK is only $110,000, slightly higher in Germany at $120,000, and less than $110,000 in France. This significant disparity in earnings makes it difficult for European companies to attract and retain top AI talent. The allure of higher salaries and more lucrative opportunities in the US is a major driver of the brain drain.
Recognizing this talent gap, the US has taken steps to attract AI professionals. In 2023, President Biden signed an executive order easing immigration rules and expanding visa categories for experts in AI and emerging technologies, making it easier for AI professionals to obtain work visas or green cards in the US. This proactive approach further exacerbates the talent shortage in Europe, as skilled professionals are drawn to the US by the promise of better opportunities and more streamlined immigration processes.
Despite the perception that Europeans prioritize leisure and high social benefits, many European IT professionals are willing to trade longer vacations for significantly higher salaries. The choice between driving a luxury car and living in a mansion on the US West Coast, flying first class, or staying in Europe and worrying about daily expenses is not difficult for many. Online forums are filled with stories of European engineers voting with their feet. The reality is that while quality of life is important, financial security and career advancement are also major considerations for many AI professionals. The significant salary gap, coupled with the perception of greater opportunities for innovation and impact in the US, is a powerful draw for European talent.
The Need for a Unifying Force
Ultimately, Europe’s AI struggles may stem from the absence of a unifying force. Although the EU has a population of 500 million and an economy comparable to the US, the European market is fragmented. The EU member states and the UK have significant differences in language, writing, and culture. The EU has 24 official languages. Companies must navigate each market individually, making it difficult to scale quickly. American tech giants can quickly dominate the market before European companies can establish a foothold. The lack of a single, unified market creates significant challenges for European AI companies seeking to expand and compete on a global scale. The costs associated with adapting to different languages, regulations, and cultural norms can be prohibitive, especially for smaller startups.
For modern large language models, robust computing power and unified datasets are crucial. While funding can address computing power, acquiring unified, high-quality datasets is a more significant challenge. The fragmented nature of European data, combined with varying privacy regulations across member states, makes it difficult to create the large, comprehensive datasets needed to train state-of-the-art AI models. This data fragmentation puts European AI companies at a significant disadvantage compared to their counterparts in the US and China, where data is more readily available and accessible.
In essence, Europe’s lagging position in the AI revolution mirrors its experience in the internet age. The lack of a unified vision, combined with structural challenges related to investment, talent retention, and market fragmentation, has hindered its ability to compete effectively in the global AI landscape. A coordinated effort, involving governments, industry, and academia, is needed to address these challenges and unlock Europe’s full AI potential.
Initiatives and Investments
The European governments recognize these challenges and have launched various AI initiatives. The EU AI Champions Initiative aims to accelerate AI development by focusing on large enterprises leading the charge. The Horizon Europe program allocates €1 billion annually to AI research and development, supporting AI development and deployment. Starting this year, an additional €1.3 billion will be earmarked for large language models and talent pool development. The InvestAI initiative seeks to raise €200 billion for further AI investment. The EU AI Act even eases regulations for small and medium-sized enterprises. These initiatives demonstrate a commitment to fostering AI innovation and addressing the investment gap.
However, these efforts might be insufficient to overcome the deep-seated structural challenges. A unifying force may be needed to truly unleash Europe’s AI potential. While increased funding and targeted initiatives are important, they need to be coupled with broader reforms aimed at creating a more supportive ecosystem for AI startups, attracting and retaining top talent, and fostering a more unified and integrated European market. The success of these initiatives will depend on effective collaboration between member states, a willingness to address the underlying structural challenges, and a commitment to fostering a long-term vision for AI development in Europe.