AI0-001 AI Implementation and Operations Practice Question
A company uses a large language model (LLM) to generate customer support responses. They notice the model sometimes produces harmful outputs. Which implementation strategy best reduces this risk while maintaining performance?
⚠ Common exam trap
CompTIA often tests the misconception that simple output filtering or prompt engineering is sufficient for safety, when in fact only training-based alignment methods like RLHF can meaningfully change model behavior without sacrificing performance.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Fine-tune the model using reinforcement learning from human feedback
Reinforcement learning from human feedback (RLHF) directly trains the model to align its outputs with human preferences for safety and helpfulness, reducing harmful outputs while preserving performance. Unlike superficial filters or prompts, RLHF adjusts the model's internal behavior through reward modeling and policy optimization, making it the most effective strategy for sustained safety improvements.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Implement a keyword-based output filter
Why it's wrong here
Keyword filters only catch literal banned strings, so paraphrased or novel harmful content passes through unfiltered, leaving the risk largely intact. They are tempting because they are cheap and fast for blocking known profanity or fixed terms, and would suit a narrow, predictable vocabulary where harmful outputs are enumerable in advance.
- ✗
Use a smaller, less capable model
Why it's wrong here
Swapping to a smaller model reduces capability across the board rather than targeting harmful content, so it degrades response quality without reliably suppressing unsafe outputs. It is tempting because smaller models are cheaper and faster to serve, which suits latency- or cost-constrained deployments where raw capability matters less than throughput.
- ✗
Add system prompts instructing the model to be safe
Why it's wrong here
System prompts are soft guidance the model can be talked around by adversarial user input, so they do not reliably prevent harmful outputs. They are tempting because they need no retraining or extra infrastructure, and would be the correct choice for steering tone, style or persona in low-risk applications.
- ✓
Fine-tune the model using reinforcement learning from human feedback
Why this is correct
Reinforcement learning from human feedback trains a reward model on human preference rankings, then optimises the LLM against it, directly suppressing harmful generations while preserving fluency and task performance. This targets the harmful-output risk at the alignment layer rather than filtering responses after generation.
About these practice questions
This AI0-001 question is part of Courseiva's 962-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →
JA
Written by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
This AI0-001 practice question is part of Courseiva's free CompTIA certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AI0-001 exam.