Sponsor
Scott Franklin
Representative · R-FL
Official title: To direct the use of artificial intelligence by National Oceanic and Atmospheric Administration to adapt to extreme weather, and for other purposes.
Introduced April 9, 2025 by Scott Franklin · Last progress April 9, 2025
Representative · R-FL
The bill would significantly improve weather forecasting, wildfire detection, research collaboration, and transparency through AI, open data, and interagency coordination — but doing so raises substantial privacy, validation, cost, IP, and capacity risks that must be managed through funding, clear governance, and strong safeguards.
Local and state governments, emergency managers, first responders, and communities (urban and rural) get earlier, faster, and more accurate weather and wildfire forecasts and impact‑based decision support, improving preparedness and reducing harm from storms, fires, floods, and other extremes.
Scientists, researchers, universities, and federal forecasters gain clearer standards, curated/open datasets, common reforecast frameworks, standardized training data, and open‑licensed NOAA data/code, enabling faster model development, reproducibility, collaboration, and operational adoption of AI tools.
NOAA, federal partners, and state/local officials benefit from mandated interagency coordination, technical guidance, and pooled expertise (including NIST, NASA, NSF, DOE, academia, and industry), improving transparency, model evaluation, and operational integration.
All taxpayers face greater federal spending pressure if appropriations are provided to build datasets, contracts, cooperative institutes, and AI systems — but if appropriations are not provided the initiatives may be limited or slow to deliver benefits.
Individuals, communities, researchers, and local businesses face heightened privacy and data‑governance risks from extensive collection, sharing, and use of observational and synthetic datasets and from integrating non‑Federal or proprietary models.
Communities and emergency responders risk harm if decision makers overrely on inadequately validated or proprietary 'black‑box' AI outputs — producing false positives, false negatives, or biased forecasts that degrade safety decisions.
Based on analysis of 16 sections of legislative text.
Authorizes NOAA-led creation of AI-ready weather datasets, evaluation of AI weather models, wildfire AI programs, workforce development, and open sharing of resulting data/code with limited exceptions.
Creates a multi-agency NOAA-led program to develop, curate, evaluate, and share large training datasets, models, tools, best practices, and workforce support for using artificial intelligence (AI) in weather, water, wildfire, and space‑weather forecasting. It directs NOAA to inventory and assess non‑Federal AI weather models, run comparative reforecast analyses with numerical models, support operational integration at forecast offices, develop an AI wildfire forecasting program, enable flexible public‑private partnerships (including co‑investment and shared IP), require open licensing for most data and code produced, and report to Congress on progress and feasibility timelines.