The electricity system was designed for a predictable world: stable generation, known demand, one-directional flows. That world no longer exists. Intermittent renewables, electric vehicles, volatile consumption and, now, data centres drawing power at an unprecedented scale. Managing that complexity with traditional tools alone is becoming increasingly difficult and costly. 

This is why AI in the energy sector has moved beyond being an add-on and is becoming part of the system’s nervous system, enabling greater capacity for analysis, prediction and response. The grid no longer simply carries energy: it is beginning to think. 

A two-way relationship 

The International Energy Agency (IEA), in its Energy and AI report, describes a two-way relationship. On one side, AI drives electricity demand upward: data centre consumption grew 17% in 2025, well above global electricity growth. On the other, AI also helps operate the system better: it enables the monitoring of grids and transformers, anticipates potential failures, improves demand forecasting and makes more efficient use of existing capacity. 

Latin America is already taking concrete steps. In Brazil, Eletrobras, one of the largest power companies in the region, is advancing the adoption of AI solutions to strengthen asset monitoring and accelerate fault detection and response. This is not an isolated experiment: it signals where regional energy operations are heading. 

Cybersecurity and governance: a condition of operation 

The same IEA report, however, raises a warning the region cannot ignore: AI also widens security risks. Cyberattacks against energy companies have tripled over the past four years, and some have become more sophisticated through the use of artificial intelligence. In critical infrastructure, digital trust is not an accessory: it is a condition of operation. 

This connects to a point that CINTEL, the Centre for Innovation, Productivity and Technological Development, has consistently underlined: integrating AI effectively requires clear rules on how data is used and how automated decisions are made. In its work on artificial intelligence and data governance, CINTEL argues that adequate governance is essential to build trust and enable the responsible scaling of AI, particularly in sectors where a single failure can disrupt essential services and affect millions of people. 

Unlocking the true potential of AI in the energy sector requires simultaneous progress on grid modernisation, data quality and availability, secure integration between operational and information technology environments, the development of specialised talent, and the adoption of robust cybersecurity and governance frameworks. 

The debate ANDICOM 2026 puts on the table 

That intersection of efficiency, critical infrastructure and digital trust is precisely the kind of discussion ANDICOM 2026, Latin America’s leading technology and artificial intelligence congress, puts on the table. The academic agenda, now available, confirms it: the CINTEL AI Forum opens with exactly that question, how to move from AI pilots to real value with responsible governance, and features a panel dedicated to how much governance AI actually needs, built on the work CINTEL has been leading in the field. Under the theme Unleashing the Power of AI, the congress will bring together leaders from the energy, technology, cybersecurity and public policy sectors in Cartagena de Indias, from 1 to 4 September 2026, to examine how to unleash the power of AI across the systems that keep an entire society running. 

Does your organisation already have an answer to that debate? 

We invite you to explore the academic agenda here. 

Sources 

• IEA – Energy and AI (2025) 

• IEA – Key Questions on Energy and AI (2026) 

• Inter-American Dialogue – Can AI Improve Latin America’s Power Grids? 

• CINTEL – Artificial Intelligence and Data Governance (May 2026) 

• ANDICOM 2026 – Academic Agenda