Artificial Intelligence applied to the industrial and enterprise world is undergoing a phase of profound transition. Now that the initial enthusiasm driven by the hype of Large Language Models (LLMs) and big technological promises has cooled down, companies and institutions are facing a reality check: to fully unlock AI's potential, a solid infrastructural foundation is required. As highlighted by Ludovico Basili – Director of Data AI & IoT at Sirti Digital Solutions – in a recent in-depth analysis on Agenda Digitale, the game of innovation and productivity will not be played on the complexity of algorithms, but on the solidity and scalability of the underlying architectures. The golden rule of this new paradigm is clear: whoever controls the infrastructure, controls the intelligence. The Algorithm Trap and the Primacy of Infrastructure Until now, the attention of the media and markets has focused almost exclusively on Artificial Intelligence models. However, in the industrial and enterprise context, the algorithm represents only the tip of the iceberg. AI is not a standalone technology, but the result of the convergence of an entire ecosystem: ultra-fast connectivity (5G Advanced), Edge Computing, flexible Cloud solutions, and IoT. Industrial AI cannot afford the latencies or approximations typical of consumer applications; it requires determinism, absolute security, real-time processing, and operational continuity. Without a network capable of transporting massive data flows and without Edge nodes capable of processing them in the field (close to machinery and production processes), even the most sophisticated predictive model remains unusable. To move from Proof of Concepts (PoCs) to large-scale production, the priority must be the modernization of network infrastructure. Data Sovereignty, Security, and the Geopolitical Challenge Another crucial theme emerging from the analysis is data governance and protection. In the era of AI, corporate data is the most valuable asset. Feeding commercial or unprotected AI models means risking the exposure of intellectual property and corporate know-how. It is therefore fundamental to invest in architectures that guarantee data sovereignty, protection from cyber threats, and compliance with regulations (such as the newly established European AI Act). Europe, aware of the computational gap compared to giants like the US and China, is accelerating infrastructural investments to ensure that businesses can innovate within a secure, reliable, and localized perimeter. The Importance of Public-Private Synergy and Skills Technological evolution cannot be sustained by the efforts of individual companies alone. It requires strong cooperation between the public and private sectors: government plans, digitalization funds (such as the PNRR and the Transition 5.0 plan), and national strategies must focus not only on the adoption of "smart" software, but on the financing of communication networks and European data centers. Finally, technological infrastructure cannot ignore human infrastructure. Introducing AI into factories and corporate processes requires a profound transformation of skills. Workforce upskilling and reskilling programs are the only way to ensure that people are capable of governing these complex architectures, transforming technology into real operational efficiency, sustainability, and new value-added services. Towards a New Industrial Paradigm Companies that successfully combine Artificial Intelligence with evolved and resilient digital infrastructures will be the only ones capable of accelerating decision-making processes and structurally increasing productivity. The true revolution of Industry 4.0 does not lie in a single miraculous algorithm, but in the ability to make platforms, data, and skills communicate within a coherent architecture. It is time to make fewer promises and build more infrastructure. Link to read the full article: https://www.agendadigitale.eu/industry-4-0/lai-industriale-ha-bisogno-di-meno-promesse-e-piu-infrastruttura/

Artificial Intelligence is not merely introducing new threats; it is radically redefining the speed and nature of digital defense. In a landscape where attacks and countermeasures move at 'machine speed,' simple reactivity is no longer enough. This profound paradigm shift was outlined by Massimiliano De Carolis, CEO of Sirti Digital Solutions S.p.A., in a recent interview with ANSA. When threats are generated and executed by advanced algorithms, traditional incident management timelines become obsolete. To protect critical infrastructure and corporate data, we must build systems capable of seeing, understanding, and acting by integrating data, technology, and context. The evolution of the cyber attack economy is severely testing conventional defense models, making it necessary to move past the approaches of the past. Among the most urgent challenges is the need to overcome the fragmentation of security tools. For years, organizations have layered isolated vertical solutions, but today it is essential to unify technologies to gain global visibility. Added to this is the rise of AI agents—powerful tools that require constant monitoring to prevent them from becoming new vulnerability vectors. Consequently, the Zero Trust model must evolve: it is no longer enough to verify human user access; this approach must extend to machine identities, constantly monitoring devices, APIs, and automated processes. In this scenario, cyber resilience takes center stage, understood as an organization's ability not only to repel an attack, but to resist, limit damage, and ensure business continuity. This profound transformation inevitably leads to rethinking the metrics by which we evaluate the effectiveness of corporate security. If until yesterday the fundamental parameters were Time to Detect (the time to identify a threat) and Time to Respond, machine speed now demands the adoption of new indicators. Time to Understand—which measures the ability to instantly correlate data to comprehend the true nature of an attack—and Time to Safe Action—the speed with which an effective corrective action can be taken without risking the business—have become crucial. Detecting an anomaly in a few milliseconds is of little use if it then takes hours to understand its context and authorize a secure response. Despite moving toward a highly automated ecosystem, Artificial Intelligence should be understood as an extraordinary force multiplier rather than an autonomous solution. Technological integration must maintain a solid governance framework at its core, where responsibility and control remain firmly anchored to the human element. Technology analyzes and accelerates, but human oversight ensures that strategic decisions are correct. Because when both attackers and defenders operate at machine speed, true competition is no longer fought on mere reaction capacity. The real competitive advantage, and the highest form of protection for future businesses, will not be reacting faster, but seeing first. The full interview with Massimiliano De Carolis: https://www.ansa.it/ansa2030/notizie/task_force_italia/2026/09/28/cybersecurity-de-carolis-con-lai-il-vero-vantaggio-sara-vedere-prima_b795b894-c4cc-4141-8978-7323b3fb4185.html