An organization uses an AI tool to scan social media for product reviews. Fraudulent social media accounts begin posting negative reviews attacking the organization's product. Which type of AI attack

The correct answer is C. Availability attack. An availability attack targets the usefulness and reliability of an AI system by corrupting its inputs at inference time, rendering its outputs unreliable or unusable. In this scenario, the AI tool is not being retrained-it is actively scanning live data. Flooding it with fraudul

AI Security Operations and Monitoring

Question

An organization uses an AI tool to scan social media for product reviews. Fraudulent social media accounts begin posting negative reviews attacking the organization's product. Which type of AI attack is MOST likely to have occurred?

Options

AModel inversion
BDeepfake
CAvailability attack
DData poisoning

How the community answered

(55 responses)

A
5% (3)
B
9% (5)
C
82% (45)
D
4% (2)

Explanation

An availability attack targets the usefulness and reliability of an AI system by corrupting its inputs at inference time, rendering its outputs unreliable or unusable. In this scenario, the AI tool is not being retrained-it is actively scanning live data. Flooding it with fraudulent reviews degrades the availability of accurate sentiment analysis, making the tool's results untrustworthy. Data poisoning (D) targets training data to corrupt future model behavior. Model inversion (A) reconstructs training data. Deepfake (B) involves synthetic media generation-none of these match the described scenario.

Topics

#AI attacks#Availability attacks#Social media analytics#Fraudulent data

Community Discussion

No community discussion yet for this question.

Full AAISM Practice