AI enters the heart of enterprises, from technology becomes critical infrastructure

(Il Sole 24 Ore Radiocor) - "The training of artificial intelligence is the largest and fastest growing business in human history, more than the advent of the railways and the industrial revolution. It is a revolution that will have more disruptive effects than the Internet'. The words spoken by Larry Ellison in Las Vegas have the flavour of a prophecy but in reality they mark the entrance into a new era that is already upon us. In fact, the AI World 2025 event was much more than a usual product launch for Oracle: it was the announcement of a paradigm shift, a strategic turning point that aims to integrate AI into every level of the enterprise architecture: from data management - now more crucial than ever to train AI - to applications, to everyday operational processes. It was also the manifesto of an accomplished transformation. The company, historically synonymous with databases, showed that it has completed its metamorphosis into an integrated and entirely AI-driven ecosystem, where artificial intelligence is no longer an add-on, but the very engine of business innovation.

At the heart of the strategy is the new Oracle AI Data Platform, described by top management as the backbone of artificial intelligence for the modern enterprise. This platform allows data to be 'AI-ready' - that is, already prepared for training, inference and automation - without intermediate steps or complex integration processes. Thanks to a unified architecture, data from different environments (public cloud, on-premise or hybrid systems) are orchestrated in real time, enabling AI to act dynamically, predictively and across business departments. This is basically the Oracle vision summarised at the event as 'AI for Data', i.e. bringing AI to where the data resides, and not vice versa. In parallel, in fact, Oracle presented Oracle Database 26ai, the new generation of its historic database, completely rethought in an AI key. Thanks to the new version, companies will be able to build and query artificial intelligence systems directly within the database, without having to transfer or duplicate data, with an enormous gain in speed, security (also because every duplication and movement of data entails a greater risk of intrusion) and operational savings. Another key novelty concerns the 'proliferation' of AI Agents - the buzzword of the moment -, today reaching the respectable figure of 400, but soon to be 600, which in Oracle's case are already integrated and made available at no additional cost within the suite of Oracle Fusion Applications, the cloud-based management software covering administration, finance and control, human resources, supply chain and customer experience processes. These intelligent agents - developed on the basis of corporate roles and tasks, and customisable by sector with a tool already integrated in the suite, AI Agent Studio - can autonomously manage processes such as expense approval, account reconciliation, candidate selection or complex order processing. Oracle explained that each AI agent can learn from user interactions and adapt to the business context, becoming a true in-house 'operational assistant'. Also because, thanks to Oracle's approach of a unified data model between databases and applications, AI agents, initially made available with training based on public data, can then be trained directly on the company's private data, preserving their privacy - which can be a winning competitive factor.

"The human being with the use of AI will replace the one without the use of AI," explained Andrea Sinopoli, Oracle's Vice President and Country Leader Cloud Tech for Italy. That is, artificial intelligence will reinforce the human component with the aim of making life better and business more efficient. But the innovations do not stop there. Oracle has also announced an expansion of multi-cloud capabilities, the result of collaboration with Microsoft Azure, Google Cloud and AWS, to allow companies to run AI workloads wherever their data resides, whether already on OCI (Oracle Cloud Infrastructure) or in the cloud of other hyperscalers. And to facilitate this, it has introduced Multicloud Universal Credits, a kind of monet expendable across all clouds for data management with Oracle.

The native integration of AI in business functions then allows repetitive processes to be automated, freeing up time and resources for higher value-added activities. In finance departments, for example, AI agents can analyse spending patterns and prevent accounting errors; in human resources, suggest career paths or optimise recruiting processes; in logistics, anticipate supply chain disruptions and propose operational alternatives in seconds. Looking ahead, Oracle's bet is to make the use of artificial intelligence accessible not only to IT teams, but to everyone in the company. Thanks to natural interfaces and simple language commands, AI becomes an everyday, transversal tool that accompanies decisions rather than complicating them. For companies, this means more efficiency, less complexity and more real-time decision-making.

As one analyst pointed out, 'Oracle is no longer talking about AI as a technology, but as the infrastructure of modern work'. A phrase that perfectly sums up the scope of this evolution: AI is no longer a goal to be achieved, but a starting condition to compete. Ultimately, the message coming out of Las Vegas is clear: companies that want to remain competitive will have to rethink their processes not according to the software, but to the intelligence that drives it. For Italy, where the entrepreneurial fabric is rich in small and medium-sized enterprises, this AI revolution - already incorporated in business and data management processes and ready to be used, according to Oracle - can represent an extraordinary opportunity for simplification and leap in scale.