CompTIA A+ Core 2 (220-1202), Domain 4, Objective 4.10 focuses on the basics of AI, and no part of that objective is more important than understanding AI's limitations and risks. AI-enabled tools can be genuinely helpful, but they can also be confidently wrong, subtly biased, and unreliable in ways that a busy technician might miss. The exam expects you to know that AI output is not automatically trustworthy and that verification by a knowledgeable human is required before acting on it. This article covers the main limitations bias, hallucinations, and accuracy or reliability limits and the verification mindset that manages them.
In this article you'll learn what AI hallucinations are and why they happen, how bias enters AI systems and their output, the practical limits on accuracy and reliability, the security and privacy risks that accompany AI tools, and a repeatable approach to verifying AI output in a support setting.