World Environment Day 2025: Can AI Fix the Energy Problem It Helped Create? "Without skilled workers to maintain these systems, long-term gains are harder to secure. AI offers clear efficiency boosts, but it's not a plug-and-play solution; it requires system-wide adaptation," says Pratik Mandvia, Solar Business Head, Mufin Green Finance

The enormous energy appetite behind artificial intelligence (AI) loudly raises concerns about the massive electricity consumption powering the everyday AI-driven innovations dominating headlines.

According to the International Energy Agency (IEA), global electricity demand from data centres could double by 2030 to around 945 terawatt-hours (TWh), slightly more than Japan's entire electricity consumption today. That's both shocking and real. But experts believe AI might be both the problem and the solution.

AI's Play in Solar

India's solar capacity has grown exponentially, with a 3,450 per cent increase over the past decade from just over 2.82 GW in 2014 to nearly 100 GW in 2025, according to the Ministry of New and Renewable Energy. The government's goal is to reach 500 GW by 2030, contributing significantly to its broader non-fossil energy target.

According to Pratik Mandvia, Solar Business Head, Mufin Green Finance, "AI in solar now goes beyond weather forecasting. It helps optimise panel tilt, detect faults, predict shading, and manage grid loads in real time."

He added that AI-led project planning has slashed construction timelines by 30 per cent and improved energy yield predictions by as much as 30 per cent. "However, limitations remain. Data quality and access are inconsistent, particularly in rural or legacy systems. Also, integrating AI into outdated infrastructure is also challenging," Mandvia explained.

Meanwhile, Anand Jain, Founder of Aerem, believes that AI use cases within solar are still developing such as Energy Management Systems (EMS) and predictive maintenance solutions. However, he noted, "The quality of data and consistency remain major hurdles."

Startups are the solutions

Despite solar's clean image, environmental trade-offs are starting to emerge. Jain said that India already generated 100 kilotonnes of solar waste in 2024, a number expected to grow to 600 kilotonnes by 2030 and a staggering 19 million tonnes by 2050. Much of this is due to the absence of a proper end-of-life (EoL) policy for solar panels.

Adding to this, Subhradeep Das, Development & Policy Economist, noted, "India must create a regulatory framework for the recycling and disposal of PV waste, similar to the EU's WEEE directive. Otherwise, we risk trading one crisis for another."

He also warned that much of the upstream manufacturing is heavily dependent on Chinese polysilicon, embedding both carbon and geopolitical vulnerabilities into the supply chain.

Mandvia echoed this sentiment and believes startups and small manufacturers can help in the long run. "India imports over 60 per cent of its solar modules from China, creating vulnerabilities in pricing and supply chains. While this has helped rapidly scale capacity, long-term resilience depends on local manufacturing. The government's PLI schemes and customs duties aim to boost domestic production. Startups and small manufacturers are gradually filling gaps, especially in modules, inverters, and battery packs," he said.

But he concludes on the note of the skilled workforce challenge. AI could become a major game-changer, but it requires widespread adoption. "Without skilled workers to maintain these systems, long-term gains are harder to secure. AI offers clear efficiency boosts, but it's not a plug-and-play solution; it requires system-wide adaptation."