A social media company trains a content moderation model. An attacker injects a small number of images into the training set that, when encountered, cause the model to classify hate speech as acceptable. The model otherwise performs normally. Which AI-specific threat is this?

ISACAAdvanced in AI Audit

Domain 2Objective 6

Threats and Vulnerabilities Specific to AI AAIA Practice Questions (Page 7)

Part of the AI Operations domain, which accounts for 46% of the AAIA exam.

31questions here

7free pages

8concepts

46%of the exam

Questions 31–31

31application · medium
A social media company trains a content moderation model. An attacker injects a small number of images into the training set that, when encountered, cause the model to classify hate speech as acceptable. The model otherwise performs normally. Which AI-specific threat is this?
Select an answer first

Finished these 1 questions?

Review the revealed explanations, or continue through the curriculum.

No more pagesBack to AAIA

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by ISACA. “AAIA” is a trademark of its owner, used for identification only.

Threats and Vulnerabilities Specific to AI AAIA Practice Questions (Page 7)

Questions 31–31

A social media company trains a content moderation model. An attacker injects a small number of images into the training set that, when encountered, cause the model to classify hate speech as acceptable. The model otherwise performs normally. Which AI-specific threat is this?

Keep studying — AI Operations