Italy's data protection chief has issued a stark warning: artificial intelligence is no longer just a technology race—it's become the defining arena of geopolitical competition, a "new cold war" where algorithms shape military deterrence, economic sovereignty, and democratic integrity. The assessment, delivered in the Italian Privacy Authority's annual report to Parliament, underscores the urgency for both regulation and investment as Europe risks being squeezed between U.S. capital dominance and China's state-driven AI push.
Why This Matters
• Sovereign risk: Italy and Europe remain heavily dependent on American and Chinese AI infrastructure, from cloud services to foundational models, threatening technological autonomy.
• Military escalation: AI-powered targeting systems deployed in Ukraine and Gaza are rewriting the rules of conflict, compressing decision-making timelines and blurring accountability.
• Investment gap: U.S. tech giants plan to pour over €700B into AI in 2026 alone, while Europe's total private AI investment over the past decade barely reaches €50B.
• Regulatory milestone: Key provisions of the EU's AI Act take effect in August 2026, imposing strict compliance obligations on high-risk systems.
The New Battlefield: AI as Infrastructure of Power
Pasquale Stanzione, president of the Italian Data Protection Authority (Garante per la Protezione dei Dati Personali), framed artificial intelligence as the "new infrastructure of power" during the annual parliamentary hearing. His assessment reflects a broader anxiety rippling through European institutions: that AI supremacy determines not just economic advantage but the very definition of sovereignty in the 21st century.
"The race for technological independence and supremacy," Stanzione explained, "mirrors a new conception of sovereignty itself." He drew explicit parallels to Cold War-era deterrence, noting that algorithmic warfare—from Ukrainian battlefields to Iranian conflict zones—has fundamentally altered the strategic equilibrium that kept major powers in check for decades.
This isn't theoretical. The 2026 geopolitical landscape has crystallized around AI dominance. U.S. President Donald Trump's visit to Beijing in May 2026 was widely interpreted as the opening salvo of this technological cold war. While Washington authorized Nvidia to export advanced H200 chips to China in late 2025, Beijing has doubled down on domestic alternatives from Huawei, signaling a commitment to full supply chain autonomy.
The stakes extend beyond semiconductors. Control over the entire "tech stack"—microchips, cloud infrastructure, foundational models, and regulatory frameworks—has become the prize in a multipolar competition with no clear referee.
Europe's Dilemma: Regulate First, Innovate Later?
The European Union has positioned itself as the global standard-setter for responsible AI through its AI Act, the world's first comprehensive regulatory framework. The law entered force in August 2024, but its most stringent provisions—governing high-risk AI systems such as biometric identification, critical infrastructure management, and employment algorithms—become enforceable on 2 August 2026.
The regulation adopts a risk-tiered approach: outright bans on unacceptable uses (like social scoring by public authorities), heavy compliance burdens for high-risk applications, and transparency mandates for general-purpose models. By December 2027, biometric systems and migration control algorithms face full scrutiny.
Yet Europe's regulatory ambition collides with an uncomfortable reality: technological dependence. As of early 2026, Europe had produced just 3 foundational AI models, compared to 40 in the U.S. and 15 in China. Despite having 30% more AI professionals per capita than America, three-quarters of European AI PhDs remain overseas after graduation.
The investment chasm is staggering. In 2025 alone, U.S. private AI investment reached $285.9B, with California accounting for $218B. Amazon, Google, Microsoft, and Meta collectively plan to exceed $700B in AI infrastructure spending in 2026. By contrast, total European private investment (EU, UK, Switzerland, Norway) in 2025 was $20.9B, and venture capital flows hover around $7-8B annually versus $60-70B in the U.S.
China, meanwhile, allocated $61.8B to science and technology in 2026, a 10% year-on-year increase, with AI, quantum computing, and 6G networks as top priorities. Beijing now holds 69.7% of global AI patents and publishes 23.2% of AI research papers.
What This Means for Italy's Economic Future
Fabio Panetta, governor of the Bank of Italy (Banca d'Italia), delivered a blunt message at a European Investment Bank conference: AI adoption can boost productivity across Italy's industrial fabric, but only if backed by "patient capital" and massive investment in intangibles—research, organizational transformation, data infrastructure.
"These investments are difficult to evaluate, with uncertain and distant returns," Panetta noted. "Italy and Europe do not start from zero. We have savings, research, and skills. But too often, resources fail to combine to support ambitious projects—or even non-ambitious ones."
The implication is clear: without mobilizing domestic and European capital, AI risks becoming a lever of development for others, not for Italy. The country's R&D spending remains stuck at 1.45% of GDP, well below the EU average of 2.38% and far behind the intensity required to compete with the U.S. and China.
Italy's National AI Strategy 2024-2026, currently under revision for the 2026-2028 period, aims to position the country as a producer of AI knowledge and technology, emphasizing quality, reliability, and specialization. But without adequate financing mechanisms—venture capital, public-private partnerships, "gigafactory" computing clusters—the strategy risks remaining aspirational.
The European Commission allocated €307.3M in new funding in January 2026, largely for AI development under the Horizon Europe program. The broader InvestAI initiative promises €20B, including sovereign supercomputing infrastructure. Yet implementation lags, and Europe's total cloud infrastructure spending in 2026 is estimated at just €10.6B, a fraction of U.S. corporate outlays.
The Human Cost: From Algorithms to Accountability
Stanzione's report didn't shy away from the darker applications of AI, particularly in conflict zones. Algorithmic targeting systems like Lavender and Gospel, deployed in Gaza, analyze vast datasets to identify targets with minimal human oversight. In Ukraine, AI-enabled drones autonomously select and engage enemy combatants.
These systems compress the decision-action loop to seconds, enabling thousands of strikes in days—a destructive capacity previously unimaginable. But they also introduce what experts call "automation bias" and "moral de-skilling," where human oversight becomes perfunctory and responsibility for lethal decisions dissolves into code.
The Italian Privacy Authority chief echoed concerns raised by Pope Francis, who recently warned against the "marginalization of humanity" inherent in unchecked AI deployment. "We must prevent reducing democracy to an equation, the person to a performance metric," Stanzione said, calling for innovation to serve human dignity rather than erode it.
Domestically, the report highlighted the "devastating" impact of sexist websites and deepfake abuse, phenomena the Authority sees as insufficiently understood by perpetrators themselves. Stanzione called for "digital pedagogy," a cornerstone of the Authority's educational mission, to equip future citizens with the skills to protect their data and, by extension, their freedom.
On media ethics, he invoked recent high-profile cases—referencing the Garlasco investigation and the "family in the woods" story—to stress that press freedom must balance with personal dignity. The principle of "information essentiality," enshrined in Italian journalistic codes, serves as a bulwark against voyeuristic spectacle and trial-by-media.
The U.S. Pivot: OpenAI's 5% Offer
Across the Atlantic, a remarkable proposal is taking shape. According to the Financial Times, OpenAI—valued at $852B—has discussed ceding a 5% equity stake to the U.S. government, aiming to secure political backing and financial support from the Trump administration.
CEO Sam Altman has reportedly suggested that other leading AI firms—Anthropic, Google, Meta—do the same, pooling shares into a sovereign wealth fund modeled after Alaska's Permanent Fund, which distributes oil revenues to the state and residents.
The proposal reflects a broader U.S. strategy: embedding AI development within national security and economic policy, ensuring that the benefits—and control—of AI remain within American borders. Whether rival firms will agree remains unclear, but the move signals how tightly AI innovation is now intertwined with state power.
Impact on Residents and Investors
For Italian businesses, the AI transition demands urgent action. Adopting AI tools can enhance competitiveness, but requires upfront investment in training, data systems, and compliance with the AI Act's transparency and risk-assessment obligations. Companies deploying high-risk AI—such as HR algorithms or credit-scoring models—must conduct conformity assessments and maintain detailed documentation by August 2026.
Investors should note Europe's regulatory lead as both an opportunity and a constraint. Compliance expertise will be in high demand, and firms that navigate the AI Act successfully may gain a "trust premium" in global markets. Yet the capital intensity of AI development favors late-stage, well-funded players, disadvantaging European startups lacking access to patient, risk-tolerant capital.
Citizens face a dual reality: AI promises productivity gains and public service improvements, but also risks job displacement (the IMF estimates 40% of global jobs will be affected, rising to 60% in advanced economies), privacy erosion, and algorithmic discrimination. The Italian Privacy Authority's emphasis on digital literacy is not incidental—understanding how AI systems use personal data is becoming a foundational civic skill.
Looking Ahead: A Multipolar AI Order
The 2026 landscape suggests a world where AI governance fractures along geopolitical lines. The U.S. exports its tech stack and capital model; China builds multilateral alternatives (including a proposed World Artificial Intelligence Cooperation Organisation in Shanghai); Europe asserts regulatory leadership through the AI Act and initiatives like the Independent International Scientific Panel on AI (IISPAI), which began operations in February 2026 under UN auspices.
India is positioning itself as a champion for the Global South in AI governance, hosting the AI Impact Summit 2026. Meanwhile, the absence of binding international treaties on military AI—exempted even from the EU's stringent regulations—leaves the most dangerous applications largely unchecked.
For Italy, the path forward requires balancing three imperatives: regulatory compliance, capital mobilization, and sovereign capability-building. Panetta's call for patient finance and Stanzione's plea for human-centered innovation are two sides of the same coin. Without both, Italy risks becoming a rule-taker in an AI world shaped by others—its data harvested, its talent drained, its industries dependent on foreign infrastructure.
The cold war metaphor is apt. Like the original, this one will be won not by ideology alone, but by industrial capacity, technological depth, and the ability to align innovation with societal values. The question is whether Europe—and Italy within it—can mobilize before the window closes.
