CISA, FBI, the National Security Agency, and international partners released AI Data Security: Best Practices for Securing Data Used to Train & Operate Artificial Intelligence (AI) Systems. This guidance provides essential strategies for protecting data throughout the AI system lifecycle, from development through deployment. It outlines critical risks such as vulnerabilities in the data supply chain, maliciously altered (“poisoned”) data, and data drift, offering best practices to mitigate these threats. The document stresses the importance of data integrity and accuracy to ensure reliable AI outcomes, recommending measures such as data encryption, digital signatures, and secure storage. Aimed at organizations handling sensitive or mission-critical data, the guidance highlights the need for robust cybersecurity protocols to safeguard AI systems and maintain the security of both training data and operational models. For a detailed breakdown of the risks and recommended safeguards, refer to the joint guidance.