AI
In cybersecurity for AI systems, normative documents play a critical role in ensuring that artificial intelligence is developed, deployed, and operated securely and responsibly. Standards like ISO/IEC 24029 (AI robustness) and ISO/IEC 5338 (AI security management) provide guidance on risk assessment, model security, and safe deployment practices, while regulatory frameworks such as the EU AI Act impose legal requirements for high-risk AI systems. Agencies like ENISA and NIST also publish guidelines on AI cybersecurity, including adversarial threat mitigation and secure data handling. Following these standards and regulations helps organizations protect AI models from manipulation, preserve data integrity, and ensure compliance with emerging legal and ethical expectations.