AWS Certified AI Practitioner (AIF-C01) — Question 38

A social media company wants to use a large language model (LLM) for content moderation. The company wants to evaluate the LLM outputs for bias and potential discrimination against specific groups or individuals.
Which data source should the company use to evaluate the LLM outputs with the LEAST administrative effort?

Answer options

• A. User-generated content
• B. Moderation logs
• C. Content moderation guidelines
• D. Benchmark datasets

Correct answer: D

Explanation

The correct choice is D, Benchmark datasets, as these are specifically designed for evaluation purposes and require minimal setup. The other options, such as User-generated content, Moderation logs, and Content moderation guidelines, involve more administrative effort to gather and analyze for bias assessment.