There are politicians who pick up an issue only when it has already become unavoidable, and there are others who start working on it while it is still the domain of specialists, researchers, and insiders. In Europe’s journey on artificial intelligence, Brando Benifei very much belongs to the second category. The Italian MEP from Partito Democratico was one of the European Parliament’s two co-rapporteurs on the AI Act during the previous legislature, and in the years following the regulation’s approval, he did not abandon the dossier once the spotlight moved elsewhere: he continued working on how those rules should be implemented, refined, simplified, and embedded into a broader European industrial strategy.
This is an important detail, because the story of European artificial intelligence does not end with the passage of a law. In fact, that is precisely where it begins. Drafting general principles is relatively straightforward compared to the task that follows: turning them into technical standards, procedures that businesses can understand, competent supervisory authorities, and rules robust enough to protect citizens and consumers yet flexible enough not to become obsolete as models evolve. It is on this less spectacular and far more practical ground that Benifei continues to operate.
From the AI Act to its implementation
The contribution Benifei is best known for remains, of course, the AI Act—the first major horizontal regulatory framework on artificial intelligence adopted by a major economic bloc. The European regulation opted for a risk-based approach: not all AI is treated equally, because a photo filter and a system deployed in a high-risk setting do not carry the same consequences. This architecture, built through years of negotiations among the Parliament, Council, Commission, businesses, and civil society, sought to avoid both blanket bans and the opposing view that market forces alone would suffice to govern increasingly powerful systems.
In 2026, Parliament revisited the file through the Digital Omnibus on AI, designed to streamline aspects of its implementation. Benifei serves as shadow rapporteur in the parliamentary procedure. The European compromise postponed certain obligations related to high-risk systems to prevent companies from being required to comply with provisions before the necessary standards and tools became available, while preserving the core framework of the AI Act. Along the same path, particularly firm measures were introduced against nudification tools and the use of AI to generate child sexual abuse material.
This phase captures a defining trait of Benifei’s work: upholding the ambition of regulations does not mean treating them as untouchable. Credible tech regulation must be adaptable when implementation reveals real-world problems, without using simplification as an excuse to hollow out its objectives. This is arguably one of the most difficult balancing acts in European digital policy, as it requires resisting two temptations at once: turning every adjustment into deregulation, and defending every regulatory detail as though altering it were a betrayal of the original vision.
Artificial intelligence is not just a safety issue
In the meantime, Benifei has expanded his focus to the relationship between AI, trade, and competitiveness. In April 2026, he served as rapporteur for a parliamentary report on the opportunities and challenges of an overarching artificial intelligence strategy for EU trade. It is a significant evolution. The first phase of the European debate was understandably dominated by risks: discrimination, surveillance, transparency, safety, and high-risk systems. The second must necessarily ask how Europe can also produce technology, scale companies, build infrastructure, and leverage AI to boost productivity and competitiveness.
This is where the debate becomes far more interesting than the old clash between “regulators” and “innovators”. Europe cannot simply limit itself to setting the terms under which technologies developed elsewhere will operate. It must possess its own models, startups, data centers, computing power, research, capital, and a market integrated enough to allow its companies to scale. At the same time, chasing the United States and China by erasing the hallmarks of the European model would be a weak strategy: it would mean competing on ground chosen by others while abandoning the advantage of building trust, legal certainty, and exportable standards.
The foresight to recognize early on that rules can become power
One of the most compelling aspects of the AI Act experience is the so-called Brussels effect: when the European single market is significant enough, rules established in Brussels can shape products and procedures beyond the Union’s borders. It does not happen automatically, nor does it mean every European policy choice will be copied, but it demonstrates that regulatory capability is a form of technological power. Benifei helped lay this very groundwork at a time when artificial intelligence had not yet reached the level of political prominence triggered by the boom in generative AI.
Yet the first-mover advantage only lasts if the rules actually work. If the AI Act were to turn into an incomprehensible maze for businesses, or if enforcement were fragmented across twenty-seven different interpretations, regulatory leadership would quickly turn into a liability. This is why the work that follows adoption is perhaps less celebrated, but just as vital as the original negotiation.
Speaking with tech builders without handing over the rules
Public records from the Parliament also reveal meetings held by Benifei in 2026 with widely diverse stakeholders across the digital ecosystem: Anthropic, Meta, Reset Tech, and Cloudflare, among others, on AI governance, cybersecurity, agents, and rights. This is precisely what ought to happen when lawmakers address rapidly evolving technology: listen to companies, researchers, civic organizations, and rights holders, while retaining political responsibility for the final decision.
Regulatory quality also depends on the ability to understand what is being regulated. In artificial intelligence, this need is particularly acute because a seemingly reasonable rule can lead to unexpected technical fallout, while a technical promise from the industry may sound reassuring without genuinely being so. Building policy expertise therefore means operating at the intersection of technology and law without mistaking dialogue for delegation.
An Italian in a debate shaping Europe's industrial future
For Italy, Benifei's role carries a value that is often underestimated. Major European tech regulations directly affect Italian businesses, public administrations, universities, and citizens. Having an Italian MEP consistently embedded in the most complex AI dossiers means taking part in shaping the rules rather than commenting on them after they have already been decided.
Italian politics still tends too often to debate technology on an episodic basis: a new app, a social media controversy, an erring chatbot, or a startup raising capital. In Brussels, the discussion is inevitably more structural, spanning the single market, standards, competition, infrastructure, fundamental rights, and trade relations. Over the years, Benifei has built an expertise that allows him to navigate this complexity with continuity—and that continuity is likely his most significant political asset on the dossier.
The next challenge: moving from regulatory leadership to technological capability
Praise, of course, is meaningful only if it does not lapse into uncritical celebration. The AI Act will have to prove on the ground that it does not stifle smaller businesses; Europe still has a massive investment gap to close compared to the United States, and market fragmentation remains a stumbling block. The real litmus test for Benifei and all the key players in European digital policy will therefore be their ability to back regulations with a credible industrial policy.
Yet this is precisely where the path taken so far proves its worth. Benifei has not remained anchored to the period when the sole priority was drafting the AI Act: his work on trade, competitiveness, simplification, cybersecurity, and governance reflects a sustained effort to keep pace with technology as it evolves. In an industry where six months can reshape products, markets, and risks, institutional continuity matters enormously.
Europe chose not to wait until artificial intelligence grew too large to be governed. It is a choice open to debate in its individual details, as with any major legislative undertaking, but politically ambitious nonetheless. Brando Benifei was one of the central architects of that choice and continues to be one in its most difficult phase: proving that protecting rights, providing legal certainty for businesses, and building genuine European AI capabilities are not mutually exclusive goals. If the Union manages to convert its regulatory leadership into technological leadership, part of the credit will belong to the patient groundwork of those who, like him, began assembling this architecture before artificial intelligence became the defining political issue of our time.
