The National Security Agency’s (NSA) Artificial Intelligence Security Center (AISC) has released a Cybersecurity Information Sheet (CSI) titled “AI Data Security: Best Practices for Securing Data Used to Train & Operate AI Systems.” The document outlines essential guidelines to protect the data that powers AI systems.

Key recommendations include the use of digital signatures to verify trusted updates, tracking data provenance, and relying on secure, trusted infrastructure throughout the AI lifecycle. The CSI also stresses the need for robust data protection measures at every stage—from development and testing to deployment and operations.

According to the NSA, the data used during the development, testing, and operation of AI systems is a core component of the AI supply chain, making its protection essential to the successful creation and deployment of these technologies. As organizations deepen their reliance on AI-driven insights, safeguarding this data becomes increasingly important to preserve the accuracy, reliability, and integrity of outcomes.

“The CSI is critically relevant for organizations — especially system owners and administrators within the Department of Defense, National Security Systems and the Defense Industrial Base — that already use AI systems in their day-to-day operations and those that are seeking to integrate AI into their infrastructure. These organizations are encouraged to adopt these best practices and mitigation strategies into their mission environments to fortify their AI systems and safeguard sensitive and critical data,” NSA stated in a press release. 

The guidance was jointly issued with cybersecurity and intelligence agencies from the United States, Australia, New Zealand, and the United Kingdom, reinforcing a unified international approach to AI system security.