OpenButterfly launches to track the environmental impact of AI data centers
Understand this story
ravipatib releases the OpenButterfly project, which tracks the environmental impact of the AI ecosystem, covering 217 data center sites, of which 144 have published or estimated planned capacity data. The project labels sites in three categories—AI-dedicated, AI-capable hyperscale, and supporting infrastructure—and lists the sources and confidence levels for water stress, grid carbon intensity, and drought risk. The project states that 68 sites, with about 6.4 GW of capacity, are located in basins rated by WRI Aqueduct as having high or extremely high baseline water stress. Capacity is taken as the full planned scale published or estimated for each site, drawn from company statements, permits, or project-side estimates; canceled sites are listed but not counted toward capacity. The project states that coverage of China and Russia is less complete than other regions, as local operators rarely publish site-level data and many sites have no capacity figures. Each site lists its information sources and a confidence rating.
AI overview based on source reports · Updated 1 hr ago
Coverage timeline
Follow the coverage to explore different perspectives.
- Hacker News AIShow HN: OpenButterfly – Tracking the environmental impact of the AI ecosystem
OpenButterfly tracks the environmental impact of the AI ecosystem across 217 sites, with capacity figures for 144 drawn from company statements, permits or estimates; cancelled sites are listed but excluded from capacity. It reports that 68 sites, about 6.4 GW of capacity, sit in basins rated high or extremely high for baseline water stress by WRI Aqueduct, and assigns water stress, grid carbon intensity and drought risk from published sources with source and confidence tiers. The project states it does not attribute environmental harm to specific facilities and that community concern flags do not imply causation.
Activity for this story
Not enough continuous observations to show a trend yet.