Italy's President Sergio Mattarella has issued a stark warning that artificial intelligence risks undermining human primacy, widening social divides, and destabilizing labour markets—a message that places Italy's head of state at the centre of an intensifying global debate over how tightly governments should regulate algorithmic systems. His intervention, delivered in a message to Pope Leo XIV on the pontiff's 71st birthday, underscores that Rome is treating AI not as a distant future problem but as a present-day governance challenge requiring immediate application of the country's new national regulatory framework.
Why This Matters
• Italy's Law 132/2025 is now fully operational, requiring businesses, professionals, and public administrators to ensure human oversight of all AI decisions.
• Transparency obligations activated in August 2026 mandate that citizens must be informed whenever they interact with AI systems, including chatbots and automated decision tools.
• A 1 billion euro investment fund for AI startups and cybersecurity has been authorized, though opposition parties dispute whether structural funding is sufficient.
• Rome and Washington are negotiating a joint technological programme to be defined within 60 days, covering quantum computing, 6G, and critical infrastructure.
The Philosophical Stakes Behind Mattarella's Warning
Mattarella's message was not ceremonial. By invoking the encyclical Magnifica Humanitas—published by Pope Leo XIV on 25 May 2026—the Italian head of state anchored his concerns in a broader moral framework. The encyclical, the first of Leo XIV's pontificate, frames artificial intelligence as a potential "new Tower of Babel," warning that technology devoid of ethical guardrails could erode human dignity and responsibility.
The President's reference to widening economic disparities signals that Rome's establishment views unchecked AI deployment as a social risk, not merely a technical one. This aligns with the encyclical's central thesis: AI systems, lacking consciousness and moral agency, must never assume roles requiring genuine responsibility.
Prime Minister Giorgia Meloni echoed this stance in remarks at the recent G7 in Evian-les-Bains, calling for a "moral compass" founded on "technological humanism." Meloni identified three fault lines: the displacement of workers, the rise of autonomous AI agents, and the proliferation of deepfakes threatening digital identity.
"If the algorithm risks displacing the intellect and the machine risks replacing the worker, politics must defend human sovereignty," the Prime Minister stated, positioning her government as a defender of European social models against Silicon Valley's faster-is-better ethos.
How Law 132/2025 Changes Daily Life
Italy became the first European Union member state to enact comprehensive national legislation complementing the EU's AI Act. Law 132/2025, which entered into force on 10 October 2025, creates binding obligations across healthcare, labour, public administration, and the courts.
For residents, the most immediate impact is transparency. Since August 2026, organisations must clearly disclose when citizens are interacting with automated systems. A chatbot handling a tax query or a municipal service request must identify itself as AI, not simulate human conversation. Content generated or modified by AI—including deepfakes—must carry machine-readable markers.
The law enshrines the principle of human-in-the-loop. An AI system may support a doctor's diagnosis or a civil servant's eligibility determination, but a qualified human must make the final decision. The legislation modifies copyright law to protect works created with AI assistance while establishing differentiated protections for source materials.
Professionals—lawyers, accountants, architects—face specific obligations. They must inform clients when AI tools are used, specify their purpose, and ensure the technology supports rather than substitutes professional judgment. Mandatory training requirements aim to prevent over-reliance on algorithmic outputs.
The Governance Architecture Taking Shape
Two agencies now share regulatory responsibility. The National Cybersecurity Agency (ACN) oversees the adequacy and security of AI systems, while the Agency for Digital Italy (AgID) manages notifications and promotes safe use cases. A biennial National Strategy for AI, coordinated by the Digital Transformation Department within the Prime Minister's office, provides strategic direction.
The law authorizes significant public investment: 1 billion euro in equity and quasi-equity financing for SMEs and startups in AI, cybersecurity, and enabling technologies, channelled through a venture capital support fund.
Subsecretary for Technological Innovation Alessio Butti defends the framework as rigorous yet innovation-friendly. "Italy has equipped itself with an organic regulatory framework coherent with the AI Act, founded on transparency, security, and human responsibility," he told ANSA. "At the same time, we are creating conditions for businesses and public administrations to increase productivity and service quality."
Butti recently returned from a mission to the United States, holding discussions at the White House, State Department, and Pentagon. The agenda covered quantum computing cooperation, 6G networks, and the energy demands of AI infrastructure. Italy and the US are expected to define a joint technological programme within 60 days.
The Opposition's Counter-Argument
The Democratic Party (PD) offers a scathing critique of government claims to "technological sovereignty."
Brando Benifei, the PD European Parliament member who served as rapporteur for the EU AI Act, called Butti's rhetoric "paradoxical." The government, he argues, has made "zero structural investments in supercomputing and public infrastructure" while aligning with "Trumpian pressures against any regulation protecting citizens' rights."
Data presents a contested picture. Italy hosts Leonardo, Europe's most powerful AI-dedicated supercomputer, operational since 2022 at the Bologna Technopole. A new infrastructure, LISA, inaugurated on 11 June 2026, operates alongside it. Major industrial players like Eni have opened their HPC6 and HPC7 systems to Italy's innovation ecosystem.
Startup financing has shifted. The standard 30% IRPEF deduction for innovative startup investments lapsed on 1 January 2026, replaced by an enhanced 65% de minimis regime deduction, capped at 100,000 euro per tax period. Public financing instruments like Smart&Start and ON Nuove Imprese remain active.
First-half 2026 data shows Italian startups raised 813 million euro across 145 operations—exceeding the same period in 2025, though with fewer individual deals.
What This Means for Residents
The regulatory framework now active creates both rights and responsibilities for people living in Italy.
Citizens encountering automated systems in healthcare, taxation, or municipal services are entitled to know when AI is involved. If a decision affects you negatively—denial of benefits, classification of risk—you have the right to request human review. The law prohibits fully automated decisions with significant personal impact.
Business owners and managers must ensure compliance with transparency obligations. Staff deploying AI tools require training. Deepfakes or AI-generated content used in commercial contexts must be clearly labelled.
The political debate over "sovereignty" will continue. The opposition is correct that Italy imports most AI solutions rather than developing them domestically; the government counters that regulatory leadership and strategic investments aim to change that equation over time.
What is unambiguous: Italy has chosen a distinctive path. Rather than the hands-off approach some tech advocates prefer, Rome has codified that algorithmic systems must serve human purposes—not the reverse. The law is on the books. The agencies are in place. The test now is whether enforcement and investment match the ambition of the rhetoric.
