In the context of AI supply chain security, which risk is most directly associated with using a third-party pre-trained model from a public repository?
#91
Answer: D✅ Correct❌ Incorrect
Which threat involves an attacker querying an AI model API extensively to reconstruct a functional replica of the model?
#92
Answer: B✅ Correct❌ Incorrect
Which technique allows multiple organizations to collaboratively train a shared AI model without sharing their raw data?
#93
Answer: A✅ Correct❌ Incorrect
In the AI model lifecycle, 'data poisoning' most commonly occurs during which phase?
#94
Answer: D✅ Correct❌ Incorrect
What is the primary purpose of the tokenization step in a Generative AI pipeline?
#95
Answer: B✅ Correct❌ Incorrect
What is the primary purpose of a Model Bill of Materials (MBOM) or AI Bill of Materials (AI BOM)?
#96
Answer: C✅ Correct❌ Incorrect
What is the primary goal of AI red-teaming?
#97
Answer: D✅ Correct❌ Incorrect
Model inversion attacks aim to:
#98
Answer: D✅ Correct❌ Incorrect
Which metric is most appropriate for evaluating whether an AI model treats different demographic subgroups equally in a binary classification task?
#99
Answer: A✅ Correct❌ Incorrect
Which principle states that AI systems must be designed so their actions and decision-making processes can be understood and traced by humans?