A new international study finds that artificial intelligence development is increasingly splitting into three distinct global systems led by the United States, China, and the European Union, each shaped by different policy priorities, innovation models, and governance philosophies. The research suggests that this divergence may permanently reshape the technological landscape and complicate global cooperation on AI safety, standards, and innovation.

A newly launched scientific journal is calling for a deeper integration of artificial intelligence into environmental research, arguing that AI may be one of the most powerful tools available to address climate change, pollution, and resource sustainability.

Researchers at the University of Arizona say there may be at least twice as many vertebrate species as previously recognized. The findings suggest that global vertebrate biodiversity has been significantly underestimated, which could have consequences for conservation efforts.

A new study found that agrivoltaics can increase or reduce yields and profits depending on the crop and where such agrivoltaic systems are deployed.

Atmospheric rivers carry unfathomable amounts of water across the sky, bringing moisture to drought-stricken regions like the Western U.S. But whether a particular incoming atmospheric river storm will result in disastrous flooding has long been difficult for researchers to determine with confidence. Now, a new DRI-led study demonstrates that accounting for soil saturation levels can substantially improve our early warning of potentially destructive flooding events. 

SAN ANTONIO — March 2, 2026 — Southwest Research Institute (SwRI), on behalf of the Numerical Propulsion System Simulation (NPSS®) Consortium, has released a new version of the NPSS software, the industry-leading aerospace software package for simulating and designing propulsion systems. NPSS 3.3 adds new functionalities, data types, and communication interfaces that provide improved flexibility and software compatibility while reducing model and component development times.