ESGreen uses AI to detect climate events and risk in credit portfolios

Economia SC reports on the study in which ESGreen cross-referenced climate forecasts generated by artificial intelligence (NVIDIA Earth-2 ecosystem) with the ESGreen Score and the locations of companies already monitored on the platform, within the NVIDIA Inception program. According to the article, the April 20, 2026 forecast run indicated two waves of instability over Rio Grande do Sul; on May 1, Rosário do Sul recorded 324 mm of rain in seven hours.

The study covered ten companies in Brazil's South and Southeast regions, with five forecast runs between April 20 and May 8, and sought to identify up to 18 days in advance which companies in a credit portfolio are in the path of an extreme event. The article cites CMN Resolution 4,945, SUSEP Circular 666 and CNSP Resolution 485 among the related regulatory frameworks.