AI energy boom sparks ESG data crisis, expert warns

By Francis Allan L. Angelo

The rapid expansion of artificial intelligence is fueling not only a global energy crunch but also a hidden data infrastructure crisis, according to ESGpedia founder Benjamin Soh.

In an email interview with the Daily Guardian, Soh said the world is focusing too narrowly on generating enough clean energy — nuclear, geothermal, and other sources — to power AI, while neglecting the “equally critical challenge of data infrastructure.”

“The real test isn’t just whether a company can get clean power; it’s whether they can prove it’s clean,” Soh said.

He described this as a “data management crisis hiding inside an energy crisis,” pointing to the growing complexity of ESG (Environmental, Social, and Governance) reporting demands faced by tech giants.

According to Soh, clean energy sources that offer “reliable, 24/7, carbon-free baseload power” — particularly nuclear and geothermal — are best suited to meet AI’s nonstop computational needs.

While solar and wind remain vital to the broader energy transition, their intermittent nature makes them ill-equipped to power AI data centers alone, he added.

Fuel cells and long-duration energy storage technologies are also key to grid stability and uninterrupted power, especially as the global reliance on AI accelerates.

Soh outlined a three-pronged ESG data challenge for major tech companies: massive data complexity, increased risk of greenwashing, and rising stakeholder pressure.

He explained that tech firms must now consolidate energy usage data from multiple sources — grid, solar, nuclear — each with different carbon intensities and data trails.

“Consolidating this into a single, accurate, and coherent sustainability report is a massive data challenge that is manual and difficult to manage with spreadsheets,” he said.

He warned that claiming “100% renewable” energy based on contracts like power purchase agreements is no longer credible without transparent verification.

To avoid greenwashing, Soh stressed that companies must implement digital ESG data infrastructure that allows real-time data ingestion, time-stamped carbon calculations, and immutable audit trails.

This digital infrastructure is essential, he said, for creating “investment-grade” ESG data — a growing requirement from regulators, investors, and the public.

Soh described the formula for success in green tech during the AI era as “Power + Proof,” arguing that verifiable data has become more valuable than the energy itself.

“In an era of intense scrutiny and mandatory disclosure, an unverified claim of ‘green’ energy has little value and carries significant reputational risk,” he said.

Investors, he noted, are now looking for hourly energy usage data, third-party assurance of emissions claims, Scope 3 supply chain carbon footprints, and metrics tracking climate transition plans.

This shift is creating new market openings for Asian technology firms, particularly those helping build ESG data platforms and decarbonize the supply chains of AI hardware.

“As global tech giants build out their AI infrastructure in Asia, they will need sophisticated local partners to help them manage their ESG data and compliance,” Soh said.

He added that Asia’s position at the center of global electronics manufacturing offers local startups a unique edge in understanding operational realities that Western firms may overlook.

However, Asian companies still face structural challenges, including fragmented national regulations and inconsistent ESG data quality across borders.

“The foundational work of data collection and cleansing is often more difficult here than in Western markets where disclosure has been common for longer,” he said.

Ultimately, the biggest roadblock to sustainable AI is the lack of harmonized ESG data standards across the global supply chain, Soh said.

“Even with the best digital platforms, the process breaks down because there is no common digital language for sustainability data,” he said.

Soh called for a universal ESG data standard — a “digital passport” — to enable seamless, real-time, and accurate carbon tracking across industries and borders.

He said achieving interoperability is essential to building scalable and efficient data systems that can support AI’s sustainable future.