Responsible AI in Eni

At Eni, we use artificial intelligence to improve operational efficiency, accelerate research and development and create value for all our stakeholders, following a human-centric approach that keeps people at the centre and in control of every tool.

Responsible AI is a way of designing, developing and using artificial intelligence systems that protects people, their rights and the quality of processes. It means defining clear rules for data, models and use cases; ensuring meaningful human oversight; assessing risks in a proportionate way; and making the purpose, functioning and limits of the tools understandable. Within this framework, technological innovation remains at the service of informed, transparent decisions and measurable benefits for all stakeholders.

Our view of artificial intelligence is clear: we adopt it in a professional and exploratory way, always respecting ethics, regulations and the central role of the individual. Every project is guided by four pillars:

• Inclusion
• Ethics
• Responsibility
• Transparency

This means preventing bias and discrimination, ensuring benefits for people and processes in line with our Code of Ethics, adopting appropriate processes and clear communication with all stakeholders, and providing timely information on how tools work, on data, sources, governance and how results are tracked. We promote a culture of responsibility in line with the European AI Act, integrated into our governance model and into widespread training programmes.

Supercomputing is the enabler of AI at Eni. HPC6, our supercomputer, launched in November 2024 at the Green Data Center in Ferrera Erbognone (Pavia), significantly increases the computing power available to our artificial intelligence tools. HPC6 enables a wide range of applications along our value chain, from 3D geosciences and computational fluid dynamics to the development of technologies to support the energy transition, and allows us to run data-intensive algorithms and workflows. Data quality is a prerequisite: we work on validated and certified datasets, developing Large Language Models and vertical models tailored to Eni’s needs, built on our data and knowledge base.

At Eni, artificial intelligence is a driver of transformation in several areas:

• operations: we apply machine learning algorithms for predictive maintenance, process optimisation and energy management, helping keep plants in optimal operating conditions and quickly identifying anomalies;
• research and development: AI accelerates the development of new materials, process optimisation and the creation of digital twins. It also supports geological and fluid-dynamics studies applied to carbon dioxide storage, innovation in batteries and simulations for magnetic confinement fusion;
• people: we simplify repetitive tasks and improve job quality through intelligent digital tools that support us in everyday activities, always ensuring human oversight.

EnergIA is the generative AI tool that makes the eni.com website more accessible and faster to use: it helps users quickly find updated content and data by querying the most relevant pages and documents published on the portal.

EnergIA is “a new window on the Eni world” available to everyone, with clear disclaimers and terms of use on AI-generated content and on the scope of the sources considered (financial documents from the last twelve months, press releases from the last two years).

We have established centres of excellence for AI governance, from prototyping to industrialisation, integrating these solutions into existing processes and investing in training and skills. Our goal is to minimise risks, preserve trust and turn innovation into more reliable services, shorter lead time and higher quality for customers and stakeholders. Putting humans at the centre and ensuring effective governance are essential conditions for generating value with AI.