Semi-supervised learning is the correct answer because it combines a small amount of labeled data with a larger amount of unlabeled data. In this scenario, humans are not fully labeling the data, but the machine is still able to correctly label current or future data points by learning patterns from the available data. That matches the concept of semi-supervised learning, where the model uses limited human-provided labels and extends learning to unlabeled examples.
Supervised learning is not the best answer because supervised learning depends on clearly labeled training data supplied by humans. Unsupervised learning is also incorrect because it identifies hidden patterns or clusters without using labels, rather than predicting correct labels for future data points. Reinforcement learning is based on rewards, penalties, actions, and an environment, which is not described here. "Semi- reinforcement learning" is not a standard main category in machine learning.
Therefore, the most accurate answer is **E. Semi-supervised learning**.

The correct answer is E. All of the above because effective prompt design requires clarity, focus, structure, and useful constraints. Information overload should not be used because giving too much unnecessary detail can confuse the model, weaken the main instruction, and reduce the quality of the response. A prompt should include relevant context, but it should avoid excessive or unrelated information.
Open-ended questions should also be avoided when the goal is a specific, controlled, or business-ready answer. Broad prompts often produce vague, incomplete, or inconsistent outputs. Instead, prompts should clearly state the desired task, format, scope, and expected outcome. Lack of constraints is also a poor prompt design practice because constraints guide the model on length, tone, structure, audience, output type, and boundaries. Without constraints, the model may generate responses that are too broad, too long, or misaligned with the user's intent.
Since information overload, overly open-ended questions, and lack of constraints can all weaken prompt quality, the correct answer is E. All of the above .

The correct answer is C. Narrow AI . Narrow AI, also called Weak AI, is designed to perform specific tasks within a limited domain. Examples include recommendation systems, chatbots, image recognition tools, fraud detection systems, voice assistants, and predictive analytics models. These systems can appear intelligent because they process data, detect patterns, make predictions, or generate responses, but they do not possess self-awareness, consciousness, emotions, independent understanding, or human-like general reasoning.
General AI and Strong AI refer to the idea of an AI system that could reason, learn, and adapt across many different tasks in a human-like way. These terms are associated with broader intelligence rather than task- specific automation. Human AI is not a standard AI category in this context. Since the question asks for the type of AI that does not have self-awareness or consciousness and operates only within defined limits, the correct choice is C. Narrow AI .

The correct answer is D. Ethics because ethics is not best treated as a single operational element of the planning and execution phase. In an AI risk framework, the planning and execution phase usually focuses on practical implementation activities such as defining the AI use case, aligning the solution with strategy, assessing financial feasibility, designing the product or solution, and preparing it for release. These activities help convert an AI concept into a working business or technical solution.
Conceptualization of the AI use case is correct because every AI initiative must begin with a clearly defined problem, objective, and intended business value. Strategy is also correct because the AI solution must align with organizational goals and risk appetite. Finance is relevant because organizations must consider cost, investment, expected return, and resource allocation. Design and release of the final product or solution is also part of execution.
Ethics is important across the entire AI lifecycle, but it is not the specific planning and execution element listed here. Therefore, the best answer is D. Ethics .

The correct answer is E. All of the above because the listed terms are associated with major functional or structural areas of the human brain often discussed when explaining cognition, intelligence, decision-making, emotions, perception, and behavior. The neocortex is linked with higher-order thinking, reasoning, language, planning, perception, and conscious decision-making. The limbic system is strongly associated with emotion, memory, motivation, and behavioral responses. The midbrain is involved in sensory processing, movement control, visual and auditory responses, and alertness. The brainstem supports essential automatic functions such as breathing, heart rate, sleep-wake regulation, and basic survival processes.
In artificial intelligence learning, these brain components are often introduced to help explain how human intelligence works and how AI attempts to model certain cognitive functions. Since neocortex, limbic, midbrain, and brainstem are all valid basic brain-related components, the best answer is E. All of the above .