Introduction

Many climate advocates view artificial intelligence as an environmental threat, picturing massive data centers devouring electricity and straining power grids. But Stéphane Germain, CEO of GHGSat, argues that framing misses a critical opportunity. In a recent Fortune commentary, he contends that climate technology cannot afford to ignore one of the most powerful tools available, especially when the challenge is turning raw satellite observations into decisions that reduce emissions.

What Happened

Germain explains that satellites already shape everyday life through weather forecasts and navigation apps, yet climate-focused satellite data remains locked behind technical complexity. Emissions, deforestation, and ice-melt observations usually require specialists before policymakers or operators can use them. At GHGSat, satellites monitor millions of facilities annually, creating far more imagery than human analysts can review in time. The company now uses AI to score which observations are most likely to show methane plumes, prioritizing them for human verification. Customers receive targeted alerts instead of overwhelming image libraries. Germain also describes broader applications, such as combining corporate disclosures, regulatory filings, and satellite measurements to track oil and gas emissions across regions, and using AI to classify deforestation alerts by likely cause. He points to geospatial foundation models—large systems pre-trained on satellite imagery—as an emerging technology that could make these capabilities routine.

Why This Matters

The stakes are high because better data only matters if it leads to action. Operators, regulators, and financial institutions need accessible emissions information to make investment, enforcement, and insurance decisions. With methane back on the agenda at COP31, the ability to measure and understand emissions without requiring a science degree could remove a long-standing barrier: ignorance. Germain is careful to note that AI is not a standalone solution to global warming, but it can translate satellite science into answers that non-scientists can use.

Key Takeaways

  • AI is often criticized for its energy footprint, but it can also support climate goals when applied to the right problems.
  • GHGSat uses AI to prioritize satellite observations likely to contain methane emissions, turning massive data volumes into actionable alerts.
  • Artificial intelligence can help merge satellite measurements with self-reported data, regulatory records, and macroeconomic context for faster analysis.
  • Geospatial foundation models are an emerging class of AI that could be adapted across many climate tasks instead of being built from scratch each time.
  • Accessible data is only the first step; decision-makers must still act on the insights AI helps uncover.

Conclusion

Germain’s message to climate tech leaders is to weigh the trade-offs of AI rather than opting out, identify a translation problem within their organizations, and ensure results reach people with the authority to act. As global negotiators prepare for COP31, the question may no longer be whether AI belongs in climate work, but how quickly its insights can move from satellite imagery to real-world decisions.