- Introduction: a meme as a starter
In some fields of socio-legal research and analysis, one of the most productive practices is to observe and listen to the expressions of ordinary people. Such observation – perhaps unsophisticated – doesn’t, however, coincide with that traditional (and largely outdated) approach of stopping passersby or neighbours to collect spontaneous opinions. Today, the role of “thermometer” of public opinion is increasingly played by digital forms of collective expression, among which memes stand out.
A striking example emerged to me during recent negotiations regarding tariffs imposed by the United States government on the European Union, when many satirical pieces of varied content began circulating on major digital platforms. By 2025, it no longer seems necessary to define a meme. Such content, whether originating from the initiative of individual users or generated by automated systems (bots) conveys perceptions and reactions to policies and regulations in an immediate way.
Among the many posts encountered – as happens to anyone with access to social networks – one was particularly suggestive. The image, grainy and set inside a bar, portrayed two men engaged in a physical fight, while a third, sitting apart and apparently indifferent, read a book wearing headphones. The scene acquired meaning through the caption: the two fighters respectively represented China and the United States competing for supremacy in Artificial Intelligence development, while the isolated reader personified the European Union, depicted as studying yet another AI document without any immediate practical outcome.
At this point, it is worth clarifying that no meme has the power (or intent) to present a fully accurate picture of reality. Like any human creation, it inevitably reflects its creator’s perspective and is neither able nor intended to provide exhaustive explanations. Memes are rather partial expressions of the digital public discourse, part of what the Packingham v. North Carolina (2017) ruling called “new digital public squares” – the entirety of the Internet and its sharing platforms.
Nevertheless, the proverbial observation remains valid: in every distorted or partial representation, some grain of truth can be found. From here arises the following reflection: is the European Union truly absent from the global race for Artificial Intelligence, or is it rather pursuing an alternative, less flashy but still significant, strategy?
- China, the United States and EU: a new “space race”
In some ways, the development of artificial intelligence could be described as the new “space race” of the 21st century. As in the last century, major world powers aim for strategic dominance over a new “territory” – now digital – to strengthen their geopolitical influence and shape a future aligned with their own principles.
Compared to the Cold War-era space race, however, there are significant cut-offs, reflecting the specificity of today’s context. First, the set of actors involved has changed: the Soviet Union no longer exists, and today’s Russian Federation is focused on political and military objectives unrelated to this technological development. Its position has been progressively taken over by the People’s Republic of China, which, under Xi Jinping’s leadership, has embarked on an accelerated process of digitalization and of closing the technological gap with the United States.
The United States, for its part, maintains a position of leadership thanks to a long historical phase of political and economic stability, which has favoured the birth and growth of the main actors of the digital revolution. The so-called U.S. “Big Tech” – including Alphabet, Amazon, Microsoft, and OpenAI – not only dominate the market, but promote a model of technological progress that imposes itself as a global benchmark; specifically, a liberalist approach in which the State acts more as a “facilitator” of private initiative than as a regulator.
A recently published macroeconomic analysis highlighted another significant point: Europe’s absence from the debate on developing advanced AI models. Considering the European Union as a single entity – like how one might view China or the United States – it is indeed apparent that it has chosen a different path. EU has not made large-scale public infrastructure investments like China, nor has it adopted a laissez-faire regulatory stance to allow private capital free rein to proliferate and create new entrepreneurial ventures.
What we said before introduces a second structural difference compared to last century’s space race: the transformation of technological governance from state-centric to hybrid. Today, private actors exert unprecedented influence, often in forms of competitive collaboration with the State. Even the Chinese model, though rooted in a rigidly centralized political structure, allows aggressive market dynamics to emerge, as shown by Huawei and DeepSeek, which embody a Darwinian “survival of the fittest” logic[1]. In the United States, this hybridization between public and private power manifests through a system of mutual legitimization: the State provides protection and strategic advantages, while tech giants help consolidate its global hegemony.
- The European Union at a Crossroad
Anyway, the figure portrayed in the meme as an isolated reader with headphones – interpreted as an allegory of the European Union – only appears inert. The European context, in fact, if examined carefully and with intellectual honesty, shows both entrepreneurial and regulatory initiatives of some significance, reflecting a peculiar approach to the global AI race.
From an industrial perspective, noteworthy examples include Mistral, a French company built on an open-source model and able to attract capital from major industries and international financial institutions. In Germany, though on a smaller scale compared to the major global players, companies like Aleph Alpha and DeepL operate, specializing in accessory applications rather than developing general-purpose AI systems.
In any case, the main criticism levelled at the EU concerns its “wait-and-see” attitude and its tendency to “over-regulate” in the tech field. Indeed, as shown by European legislative acts, codes of practice, and related sectoral studies, the Union’s priority has always been to “ensure that AI systems used in the EU were safe, transparent, traceable, prevented bias and discrimination, […] and ensured respect for fundamental rights.” In particular, the humanist principle has from the outset been emphasized as the guiding light for creating AI systems that are not dangerous to society and the individual – interests ranked even above the market.
Within this perspective falls the recent GPAI Code of Practice, complementary to the AI Act – particularly Articles 53, 55, and 56 – which introduces detailed technical standards to guarantee the safety and traceability of so-called general-purpose AI models.
However, even here, the approach may seem unassertive: the act’s voluntary adherence regime can reinforce the image of a hesitant Union faced with rapid technological evolution. In this sense, Meta’s decision – not to sign the new code of practice – is telling. As one of the most active companies in AI development, and already a frequent interlocutor of the EU in content moderation through the Media Freedom Act and the Digital Services Act, Meta’s stance raises a crucial question: will the European Union be able to assert its regulatory vision in a competitive and fast-changing global context?
- Rome Wasn’t Built in a Day
In conclusion, the question initially raised by the meme lends itself to a more complex reflection. The metaphor, though useful as an introductory prompt, is insufficient to capture the dynamics of the trilateral competition between the United States, China, and the European Union – a competition destined to decisively influence the future of artificial intelligence and, consequently, global society in the digital age.
A grain of truth, however, can be found, and on it we can base our brief final consideration.
Ultimately, the core issue is the kind of future envisioned for European AI. Taking the matter to an extreme, if one assumes a “cyberpunk” scenario as desirable – marked by rapid technological progress with little regard for citizens’ rights and democracy – the European approach is structurally disadvantaged, being slower and more costly in terms of time, energy, and financial resources.
On the other hand, a quick (as in the one provided here) comparative analysis of alternative models suggests a cautious outlook. The Chinese paradigm, as echoed in the reflections of philosopher Zhao Tingyang[2], appears to configure a post-democratic order, where fundamental rights and liberal values become residual elements of a bygone era. The U.S. model, instead, emphasizes technological leadership and market dominance by large private actors, leaving uncertain the distribution of benefits for society and the actual balance between economic power and social protection.
In this context, the European approach – though marked by supposed slowness and apparent inaction – emerges as a coherent continuation of an institutional path centred on the human being and the protection of fundamental rights. This strategy, not immediately productive of results, aligns with what scholars have termed the Brussels Effect (Bradford[3]) – the EU’s regulatory capacity to progressively extend its standards to the global context, aiming to create a fair and sustainable culture of technological progress.
While waiting for this process to fully unfold, it is essential that the Union holds firm to its regulatory vision, accepting temporary sacrifices in view of a long-term goal: ensuring reasonable, safe, and dignity-respecting development of Artificial Intelligence.
Matteo Paolanti is a PhD Candidate in Comparative Constitutional Law at the University of Siena.
[1] In Italy, an interesting essay was recently published on the condition of China with respect to the typically Western view that is often and willingly given. Although there are other technical texts useful for understanding today’s China (just think of Sujian Guo’s fundamental volume ‘Post-Mao China’), we refer you to: Yoshimi, Cose incredibili accadono in Cina (Incredible Things Happen in China), Rome, 2025.
[2] For further information on this topic, see Z. Tingyang, All under Heaven: The Tianxia System for a Possible World Order, Berkeley, 2021.
[3] A. Bradford, The Brussels effect, Oxford, 2019
