Artificial intelligence (AI) emerges as both a lifeline and a liability for the energy sector. It is a vital tool for forecasting and renewable energy integration, but also a source of new risks, including bias and inequality, as well as the staggering electricity consumption of data centers.
This was highlighted by experts of the second plenary session of the 11th Annual Public Policy Conference (APPC), organized by the Philippine Institute for Development Studies (PIDS) as part of the 23rd Development Policy Research Month (DPRM).
Dr. Marnie Shaw, Associate Professor and Deputy Director of The Australian National University, highlighted AI’s promise as an “orchestrator” for a highly distributed energy system.
In Australia, where one in three homes has rooftop solar, AI is already helping households, communities, and large-scale networks manage solar, batteries, and grid operations.
“AI is very helpful for that task because it’s so good at forecasting,” Shaw said, stressing its role in balancing electricity supply and demand.
However, she pointed out several risks, including “lack of transparency, biased or poor training data, inequality impacts, and value-insensitive decision making.”
“We don’t become so fearful of AI risks and over-governed AI that we lose sight of the opportunity to use AI to solve the greatest challenge of our time,” she added.
Addressing the ‘energy trilemma’
For Dr. Roy Francis Navea, Full Professor at De La Salle University, AI’s role must be understood within the “energy trilemma” of security, equity, and sustainability.
“Traditionally, these goals have been viewed as trade-offs, but current pressure requires simultaneous advancement in all three areas,” he explained.
Navea cited AI’s applications in forecasting, predictive maintenance, and grid optimization.
“AI efficiently manages electricity flow across complex networks and integrates distributed energy sources, contributing to a more decentralized and renewable power system,” he said.
He cited innovations such as a flywheel energy management system developed with Amber Kinetics Philippines.
However, Navea also highlighted the paradox of AI in energy. “The operation of sophisticated AI models and the data centers that support them consumes a substantial amount of energy,” he noted, echoing global concerns about rising electricity and water use.
Industry perspective and policy needs
With an industry lens, Mr. Norman Ocaña III, Government Affairs & Sustainability Leader at Schneider Electric Philippines, emphasized the speed of adoption and implications of AI.
“Each Generative AI query requires 10 times more energy than your typical Google search. Therefore, plans to harness AI for societal benefit must also account for its rising energy demands,” he stressed.
Ocaña noted the unprecedented pace of AI adoption, with generative AI reaching 100 million users in just two months compared to the internet’s seven years.
Schneider Electric has leveraged AI to optimize manufacturing facilities, resulting in reduced downtime, improved inventory forecasting, and shorter delivery times, he said.
Looking ahead, Ocaña outlined three key policy priorities: a robust renewable energy market, support for on-site power generation, and standardization of open-source power management software.
Inclusive electrification in the Philippines
Former Energy Regulatory Commission Chairperson Atty. Monalisa Dimalanta brought the discussion into the Philippine policy context, linking AI to the implementation of the E-Governance Act and to efforts to expand electrification.
“If there is AI that will allow for better management of these networks, then… they increase reliability in the system because they’re able to forecast and predict,” she said, citing its application in rooftop solar integration and spot market operations.
Dimalanta stressed AI’s potential in extending power access to underserved areas.
“We actually have 250 islands in the Philippines that are not yet connected to the [off-grid electrification] system. If we are able to introduce that flexibility of AI through hybrid systems, then we will increase the ability to electrify these islands and really promote development in a more inclusive way,” she explained.
She concluded that AI must be “assistive, accountable, and uplifting” to ensure inclusivity and sustainable development.
Catch the playback of the APPC Plenary Session 2 here: https://bit.ly/recordAPPCSession2. — RTG
