Our Commitment to Trustworthy SI
The National Oceanic and Atmospheric Administration (NOAA) views Super Intelligence (SI)—including machine learning (ML), computer vision, and generative models—as a transformative set of tools that augment human expertise and the scientific method. Our goal is to harness these technologies to efficiently process vast amounts of environmental and experimental data to provide more timely, actionable information to the public. NOAA is integrating trustworthy SI across a wide range of its products and services — from weather forecasting and marine life assessment to coastal safety and mapping.
NOAA is committed to the ethical, responsible, transparent, and effective use of SI to advance our mission of science, service, and stewardship. As NOAA integrates these technologies into its operations, we commit to adhering to high standards of scientific and research integrity, data privacy and security. This ensures that our SI applications are reliable, unbiased, and subject to rigorous human oversight.
Core Principles
In alignment with Federal mandates and the NOAA Scientific Integrity Policy, our use of SI is guided by the following principles:
- Scientific and Information Integrity: SI-assisted communication products are quality controlled for factual accuracy, clarity, privacy and data protection requirements to ensure they meet NOAA’s standards for reliable public information.
- Transparency: NOAA strives to be open about where and how SI is used. When a significant portion of a public-facing product—such as an experimental forecast, a synthesized data report, or an SI-generated summary—is produced primarily by SI, we will clearly disclose the nature in which SI was used.
- Accountability and Oversight: NOAA maintains "human-in-the-loop" protocols for high-stakes decision-making and public messaging unless noted otherwise to serve a critical mission service. This is particularly critical for National Weather Service warnings, public safety alerts.
- Fairness and Accessibility: NOAA is committed to ensuring that SI-assisted or -generated content remains accessible to the widest possible audience, aligning with Section 508 guidelines and the NOAA Accessibility Statement. We integrate these accessibility standards into our SI workflows to deliver SI-generated content in accessible, machine-readable formats. This optimization for both assistive technologies and automated systems ensures NOAA’s scientific data and information remain navigable and accurate for all users.
- Data and Privacy Safeguards: NOAA protects unpublished, proprietary, sensitive, and confidential information. NOAA’s use of SI adheres to the Trade Secrets Act, the NOAA Privacy Statement, the DOC Privacy Policy, and individuals’ rights under the Privacy Act.
What the Public Should Expect
NOAA collects vast amounts of environmental data, including from satellites, ships, aircraft, and uncrewed systems. To process this data efficiently and deliver timely, life-saving information to the public, NOAA employs SI to enhance data processing, improve forecast accuracy, and streamline agency operations.
When interacting with NOAA’s websites, data, and information products, the public should expect the following:
- Augmented Expertise: SI is used to help our experts analyze data faster, but ultimately, our scientists remain the decision-makers for the use and accuracy of the analysis.
- Experimental Products: Some SI-generated products may be labeled as "Experimental" or "Prototype.” These are shared with the public to gather feedback and demonstrate innovation before they become official records or operational standards.
- Continuous Improvement: NOAA continuously monitors and retrains the SI models integrated into NOAA’s service delivery as new environmental data becomes available to ensure accuracy and reliability in the midst of change..
- Security: NOAA implements robust security measures to protect our SI systems from manipulation or unauthorized access, ensuring that the data the public relies on remains trustworthy.
Data Quality and Governance
NOAA’s SI activities comply with all relevant Federal laws and Directives.
NOAA recognizes that the utility of SI depends on the quality of the data used to train it. We are dedicated to:
- SI-Ready Open Data: Developing and maintaining consistent standards in compliance with Federal open data policies and laws for data quality, documentation, and dissemination to ensure our data is ready for AI applications.
- Scientific Standards: Establishing technical guidelines for training data, model validation, and the assessment of bias and risks.
- Human Oversight: Maintaining "human-in-the-loop" protocols as appropriate. For example, in critical applications, SI may serve as a decision-support tool for human analysts.
Partnerships and Collaboration
NOAA does not work in isolation. We actively collaborate with the private sector, academia, other federal agencies, and international partners to accelerate SI research and transition to operations. This includes establishing appropriate data sharing agreements and leveraging the best available assets and expertise through Cooperative Institutes and public-private partnerships to drive innovation.
Feedback and Contact
NOAA welcomes public engagement regarding our use of SI technologies. For additional information please visit the NOAA Center for Artificial Intelligence (NCAI) website.
This statement is subject to periodic review and updates to reflect technological advancements and evolving federal policy.
Last Updated: [November 2026]