CompTIA AI Fundamentals Study Guide & Cheat Sheet

CompTIA AI Fundamentals Study Guide

Exam domains

Who it’s for: learners and non-technical professionals who want practical, vendor-neutral AI literacy

Study & test-day tips

• Focus on concepts over jargon: you should be able to explain in plain words what AI, machine learning, and generative AI are and how they differ, without any math or coding.
• Remember the one-line model of how chatbots work: a large language model predicts the most likely next word over and over, based on patterns it learned from huge amounts of text.
• Treat every AI answer as a confident first draft, not the truth. The exam rewards the habit of verifying facts and sources before you trust or share output.
• Learn to recognize a hallucination: an AI can state wrong information, fake citations, or made-up details in a fluent, convincing tone, so confidence is never proof of accuracy.
• Practice the anatomy of a good prompt: give clear instructions, useful context, a role or goal, the format you want, and an example when it helps.
• Iteration is a skill, not a failure. If the first answer is off, refine the prompt, add detail, or ask for changes rather than starting over.
• Always ask 'is AI the right tool here?' AI fits drafting, summarizing, and brainstorming, but is a poor fit for guaranteed accuracy, confidential data, or final human judgment.
• Know the responsible-AI risk list cold: bias, hallucinations, privacy, security, transparency, and intellectual property or copyright, and be ready to give a simple example of each.
• Protect sensitive information: never paste passwords, personal data, or confidential company material into a public AI tool, since you may not control how that data is stored or used.
• Keep tools generic. Brand names like ChatGPT, Claude, or Gemini are only examples of chatbots and image generators; the principles apply to any AI tool, which is the point of a vendor-neutral credential.

Cheat sheet

Core AI concepts

• Artificial intelligence (AI): software that performs tasks normally needing human intelligence, such as understanding language or recognizing images.
• Machine learning (ML): a part of AI where systems learn patterns from data instead of following hand-written rules.
• Generative AI: AI that creates new content such as text, images, audio, or code based on patterns it learned.
• Large language model (LLM): the engine behind chatbots; it predicts the most likely next word to produce fluent text.
• Training data: the large collection of examples a model learns patterns from; its quality and bias shape the output.
• Model vs tool: the model is the underlying brain; the chatbot or app is the friendly interface you actually use.

How chatbots and LLMs work

• An LLM works by predicting the next word (token) again and again until it forms a full response.
• It has no true understanding or beliefs; it produces statistically likely text, not verified facts.
• Output can vary between attempts, so the same prompt may give slightly different answers.
• A context window limits how much text the model can consider at once, so very long inputs may be cut off.
• The model does not automatically know recent events or your private files unless they are provided to it.

Writing effective prompts

• Be specific: state exactly what you want, for whom, and why.
• Add context: give background, audience, and any constraints the AI should respect.
• Assign a role or goal: for example, 'act as a study coach' or 'help me summarize this for a beginner.'
• Specify the format: ask for a list, table, short paragraph, or step-by-step output.
• Show an example: a sample of the style or structure you want guides better results.
• Iterate: refine, correct, or expand the prompt instead of accepting a weak first answer.

Evaluating and verifying output

• Treat AI output as a draft that must be checked, not a final answer.
• Verify facts, names, numbers, and quotes against trusted, independent sources.
• Be suspicious of citations and links the AI provides; they can be fabricated.
• Watch for hallucinations: fluent, confident text that is actually wrong or invented.
• Judge quality by relevance, accuracy, tone, and whether it truly answers your need.

Responsible and ethical AI

• Bias: AI can reflect unfair patterns from its training data, so review for fairness.
• Privacy and security: do not enter passwords, personal, or confidential data into public AI tools.
• Transparency: disclose when content is AI-assisted where honesty or rules require it.
• Intellectual property and copyright: AI output may resemble existing work, so check ownership and usage rights.
• Accountability: a human stays responsible for decisions and for checking AI output before it is used.

Applying AI and knowing its limits

• Strong fits: drafting, rewriting, summarizing, brainstorming, explaining, and studying.
• Practical uses: turning notes into summaries, simplifying complex text, and outlining plans.
• Basic automation: combine AI with simple tools to handle repetitive, low-risk tasks.
• Poor fits: tasks needing guaranteed accuracy, confidential data, or final human judgment.
• Career readiness: AI literacy is a transferable skill that boosts productivity across many roles.
• Golden rule: keep a human in the loop to review, verify, and approve important output.

Glossary

Artificial intelligence (AI)

Software that performs tasks normally requiring human intelligence, such as understanding language, recognizing images, or making predictions.

Machine learning (ML)

A branch of AI in which systems learn patterns from data and improve at a task instead of following explicitly written rules.

Deep learning

A type of machine learning that uses layered neural networks to learn complex patterns from very large amounts of data.

Neural network

A computing structure loosely inspired by the brain that learns to recognize patterns by adjusting connections during training.

Generative AI

AI that creates new content such as text, images, audio, or code based on patterns learned from existing examples.

Large language model (LLM)

An AI model trained on huge amounts of text that generates language by predicting the most likely next word.

Chatbot

A conversational application, often powered by an LLM, that responds to user messages in natural language.

Token

A small chunk of text, such as a word or part of a word, that a language model reads and predicts one at a time.

Next-token prediction

The core mechanism of an LLM, which repeatedly predicts the most likely next piece of text to build a response.

Prompt

The instruction or question a user gives to an AI tool to guide its response.

Prompt engineering

The practice of writing clear, specific, and well-structured prompts to get better results from an AI tool.

Context

The background information and details included in a prompt that help the AI produce a relevant and accurate response.

Context window

The maximum amount of text an AI model can consider at once; input beyond this limit may be ignored or cut off.

Iteration

Refining a prompt step by step, based on the AI's previous answers, to steer the output toward what you need.

Training data

The large collection of examples used to teach an AI model; its quality and biases directly affect the output.

Hallucination

When an AI produces information that sounds confident and fluent but is actually false, invented, or unsupported.

Bias

Unfair or skewed patterns in AI output that often come from imbalances or prejudices present in the training data.

Fairness

The goal of ensuring AI systems treat people and groups equitably and do not produce discriminatory outcomes.

Transparency

Being open about when and how AI is used, including disclosing AI-assisted content when honesty or rules require it.

Privacy

Protecting personal and sensitive information by not sharing it with AI tools that may store or reuse it.

Data security

Practices that keep information safe, including avoiding entering passwords or confidential data into public AI tools.

Intellectual property (IP)

Legal rights over creative or original work; AI output may raise questions about ownership and permitted use.

Copyright

A legal protection for original works; AI-generated content may resemble protected material, so usage rights must be checked.

Responsible AI

Using AI in a safe, fair, transparent, and accountable way that respects people, privacy, and the law.

Human in the loop

Keeping a person involved to review, verify, and approve AI output before it is trusted or acted on.

Verification

Checking AI output against trusted, independent sources to confirm that facts, numbers, and claims are correct.

Fact-checking

The process of confirming that statements produced by AI are accurate before relying on or sharing them.

Automation

Using technology, sometimes with AI, to handle repetitive tasks with little or no manual effort.

Image generator

A type of generative AI that creates pictures or artwork from a text description provided by the user.

Summarization

Using AI to condense long text into a shorter version that captures the key points.

AI literacy

The practical ability to understand, use, evaluate, and question AI tools responsibly in everyday work and study.

Vendor-neutral

Not tied to any single company or product, so the skills and concepts apply to any AI tool you might use.