Valid ISO-IEC-27001-Lead-Implementer Dumps shared by EduDump.com for Helping Passing ISO-IEC-27001-Lead-Implementer Exam! EduDump.com now offer the newest ISO-IEC-27001-Lead-Implementer exam dumps, the EduDump.com ISO-IEC-27001-Lead-Implementer exam questions have been updated and answers have been corrected get the newest EduDump.com ISO-IEC-27001-Lead-Implementer dumps with Test Engine here:
Question: During a security audit, analysts discover that an attacker repeatedly queried a black-box ML model to infer if specific data points were in the training set. The attacker could determine if an individual's data was used during training. What threat does this attack represent?
Correct Answer: C
ISO/IEC 23894:2023 (Artificial Intelligence Risk Management) and NIST SP 800-207A define Membership Inference Attacks (MIA) as: "An adversary attempts to determine whether specific data was used in the training phase of a machine learning model." This is a privacy threat and can lead to data breaches, especially with personally identifiable information (PII). It differs from data poisoning, which manipulates the training process, and backdoors, which alter behavior intentionally. References: ISO/IEC 23894:2023 Clause 8.2 - Machine Learning Threats ISO/IEC 27001:2022 - Controls A.8.10 and A.8.12 (Data protection, leakage prevention)===========