Questions over Italy’s positioning in the growing technological competition between the United States and China remain unresolved following the debate sparked by AgID Director General Mario Nobile’s participation in China’s World Artificial Intelligence Conference (WAIC).
As Decode39 recently reported, the episode has yet to receive an official explanation from either AgID or the Italian government, leaving what appears to be an enduring ambiguity over the institutional significance of the Shanghai visit.
Decoding the news: That ambiguity, however, reflects a much broader strategic dilemma. As competition between Washington and Beijing accelerates, governments increasingly face a difficult question: how long can they continue to engage with both AI ecosystems before strategic choices become unavoidable?
- For Beniamino Irdi, CEO of HighGround and Senior Fellow at the German Marshall Fund of the United States, artificial intelligence has already become one of the defining arenas of geopolitical competition.
A competition beyond the Cold War. “Artificial intelligence is becoming one of the main areas of strategic competition between the United States and China,” Irdi says. Yet he believes comparisons with the Cold War only go so far. While nuclear weapons represented a strategic technology concentrated in the military domain, AI is fundamentally different because it is a general-purpose technology capable of reshaping virtually every industrial and economic sector.
- “That makes the dilemma even more acute,” he argues, explaining that leadership in AI is likely to have consequences extending well beyond military power and into the broader balance of economic and technological supremacy.
- Rather than looking to the Cold War, Irdi sees a more useful parallel in China’s Belt and Road Initiative: Beijing first expanded its influence through transport infrastructure and later through digital networks, particularly across the Global South. Today, he argues, the same strategic logic is being applied to artificial intelligence.
Adoption first. “The goal is not to offer the most sophisticated product,” Irdi says. “The goal is to become the default supplier, set the standards and transform widespread adoption into political and economic leverage.”
- In Irdi’s view, this explains why China is not necessarily competing to dominate the technological frontier. Instead, Beijing is seeking to make its AI ecosystem the most accessible by promoting affordable models, open-source solutions and broad international availability. The objective is to maximise adoption first and consolidate strategic influence afterwards.
The battle for standards. According to Irdi, this is precisely why the debate over standards has become as strategically significant as the race for chips and computing power.
- “Computing capacity and semiconductors constitute the hardware of power,” he says, noting that the United States continues to enjoy a substantial lead in frontier models, advanced chips and cloud infrastructure.
- “But standards, governance frameworks and technological partnerships concern diffusion rather than restriction,” he adds. “This is the level on which Beijing chooses to compete.”
Why WAICO matters. China’s push for open-source AI models and new multilateral initiatives should therefore not be seen simply as technology diplomacy. Irdi points to the recent establishment of the World Artificial Intelligence Cooperation Organization (WAICO) in Shanghai as part of a broader effort to shape global AI governance, particularly across emerging economies.
- “The objective is to dictate the conditions, become the predefined option and transform that position into strategic leverage,” he says, arguing that the contest over AI is increasingly being fought through the diffusion of technological ecosystems rather than through frontier capabilities alone.
Why ambiguity comes at a cost. If standards increasingly define technological ecosystems, Irdi argues, governments also have less room to postpone strategic choices.
- “There is still considerable room to interact with both AI ecosystems when it comes to individual tools, commercial applications and research,” he says. “But that margin becomes very limited when decisions concern standards, public administrations and governance frameworks.”
- According to Irdi, the reason is simple: those decisions are far more difficult to reverse than the adoption of individual AI applications. Once governments, developers and suppliers have standardised around a particular technological ecosystem, the strategic choice has effectively already been made—even if no political decision has ever been formally announced.
“The danger,” he argues, “is not only that the choice is becoming binary, but that it becomes conditioned automatically by decisions that continue to be regarded as reversible.”
- For the expert, this is precisely where many governments underestimate the strategic implications of AI governance. The debate often focuses on whether countries can continue to balance relations with competing technological ecosystems. In practice, however, he argues that institutional choices gradually narrow that room for manoeuvre until the balance disappears almost by itself.
The market follows political signals. That dynamic explains what Irdi considers one of the central misunderstandings in today’s AI competition.
- “Ambiguity is not a neutral position. It is a cost,” he says.
- Governments often assume that maintaining strategic ambiguity allows them to preserve flexibility. In reality, ambiguity sends a signal to the market, which naturally gravitates towards the most affordable and immediately accessible technologies.
- “Where governments do not provide a clear strategic direction, the market automatically turns towards the most affordable and accessible option,” he explains. “Economic convenience tends to prevail in the short term over what should instead be a long-term strategic direction.”
That process, Irdi argues, can gradually generate technological dependence without any explicit political decision ever being taken. Public administrations, universities, smaller companies and local institutions are all likely to adopt the ecosystem that offers the lowest barriers to entry, making affordability itself a geopolitical advantage rather than merely a commercial one.
- The paradox, he warns, is that governments may eventually find themselves caught between two worlds.
- “If a state formally remains aligned with the frontier ecosystem but fails to provide concrete guidance to the market, the result can be the worst possible,” Irdi says. On the one hand, that country risks no longer being considered a fully reliable partner in discussions involving sensitive technologies, standards and security. On the other, it may simultaneously develop a growing dependence on an alternative ecosystem that was never consciously chosen in the first place.
A question of sovereignty. The consequences of strategic ambiguity extend well beyond governments and public administrations. They also shape the decisions of private companies, which need clarity to make long-term investments.
- “From the point of view of the industrial base, companies do not have enough information to make long-term investments if they do not know which technological stack they will have to adopt and which ecosystem they will ultimately rely on,” Irdi says. “Strategic clarity is therefore a prerequisite for industry to invest with a long-term horizon.”
- That is why the choice between competing AI ecosystems should not be treated as a purely commercial matter. “We must not make the mistake of treating this choice as a commercial decision,” he says. “It is ultimately a matter of national security and sovereignty that requires clear public direction.”
The bottom line: At the same time, Irdi acknowledges that the geopolitical environment has become more complicated for European governments. Recent signals from Washington, he argues, have introduced an additional layer of uncertainty into the transatlantic relationship, making strategic planning even more difficult.
- “That makes it even more evident that Europe needs to accelerate the development of its own sovereign ecosystem. It is a difficult objective and we are already very late, but it increasingly appears to be the only way to overcome this dilemma.”
