Responsible and Reliable AI

Ensuring AI advances Earth system science with integrity, transparency, and ethical practice.

What is Responsible and Reliable AI in Earth System Science?

Responsible AI practices ensure transparent, clear information about AI tool design, effective use, and appropriate application.

Reliable AI means AI tools that work consistently, deliver scientifically valid results, and meet rigorous standards for their intended applications.

In Earth system science, this means building AI on verified high-quality data matched to specific purposes, creating reproducible workflows, and rigorously validating performance across accuracy, physical consistency, and real-world applicability. High-quality, purpose-matched data is the foundation of reliable AI.

How can you apply responsible and reliable AI in your own research? Apply responsible and reliable AI in your research.

Why Is NSF NCAR Committed to Responsible and Reliable AI?

AI's role in enhancing our understanding of Earth system processes, from weather forecasting and hazard preparedness to long-term prediction, influences decisions that affect safety, economy, and security, and requires thorough evaluation and documentation.  

As a community research facility, NSF NCAR has a unique responsibility to establish standards and practices that shape how we develop trustworthy AI and apply those practices and applied across the broader research community.

Learn more about NSF NCAR’s priorities and commitments for responsible and reliable AI in ESS: NSF NCAR's priorities and commitments.

See our growing list of community resources and reference materials.

For more information about our responsible and reliable AI program, contact Monica Morrison.