Skip to content
Skuto

AI glossary: plain-language definitions

AI comes loaded with jargon, and most explanations assume you already know it. This glossary doesn't. Every term gets a plain-language definition first, then a short note on where you'll actually run into it.

Put it to work

Put the words to work

Definitions are a good start. These free tools let you actually use the vocabulary right away.

49 terms, growing weekly

A

  • AI agent : An AI agent doesn't just answer. It takes actions for you: browsing sites, filling forms, running multi-step…
  • AI assistant : An AI assistant is a program you talk to in everyday language that helps you write, plan, summarize, and solve…
  • AI company policy : An AI company policy is the internal document that says how staff may use AI tools at work: which tools are…
  • AI detector : An AI detector is a tool that claims to tell whether a text or image was made by AI. The honest truth: they…
  • AI memory : AI memory is a feature that lets an assistant remember facts about you across conversations: your job, your…
  • AI model : An AI model is the trained "brain" inside an AI product, the result of teaching a program patterns from huge…
  • AI watermarking : AI watermarking embeds an invisible marker in AI-generated text, images, audio or video so machines can later…
  • Anonymization : Anonymization means stripping data of anything that could identify a person, so it stops counting as personal…
  • API : An API (application programming interface) is the doorway programs use to talk to each other. "Using AI via…

C

  • Chat history : Chat history is the list of past conversations an AI assistant saves to your account. It's handy for picking…
  • Chatbot : A chatbot is any program that answers you in a chat window. Older ones followed rigid scripts; modern AI…
  • Confidential data : Confidential data is information you're obliged to keep restricted: client lists, contracts, financials…
  • Context window : The context window is how much text an AI can keep in mind at once: your messages, its replies, and any pasted…
  • Copyright and AI : Copyright and AI is the open question of who owns what: whether AI companies may train on copyrighted works…
  • Custom instructions : Custom instructions are standing notes you save in an AI's settings (who you are, how you want answers) so it…

D

  • Data breach : A data breach is when personal data ends up accessed, leaked or stolen by people who shouldn't have it. Under…
  • Data controller : Under the GDPR, the data controller is whoever decides why and how personal data is used. If your business…
  • Data protection authority : A data protection authority (DPA) is the public body in each EU country that enforces privacy law. It…
  • Data retention : Data retention is how long an AI provider keeps your conversations on its servers, including after you delete…
  • Deep research mode : Deep research mode is an AI feature that spends several minutes searching dozens of sources, cross-checking…
  • Deepfake : A deepfake is AI-generated audio, video or imagery that convincingly imitates a real person: their face, their…
  • DPA (Data Processing Agreement) : A DPA is the contract between your business and a provider that processes personal data for you, as required…

E

  • Enterprise tier : The enterprise tier is the top business plan of an AI service, built for organizations: admin controls, single…
  • EU AI Act : The EU AI Act is the European Union's law regulating artificial intelligence. It sorts AI uses by risk…

F

  • Fine-tuning : Fine-tuning is extra training that adapts an existing AI model to a specific job or style using your own…
  • Free tier : The free tier is the no-cost version of an AI service. ChatGPT, Claude, Gemini, and Copilot all have one. You…

G

  • GDPR : The GDPR (General Data Protection Regulation) is the EU law that has governed personal data since 2018. It…

H

  • Hallucination : A hallucination is when an AI states something false with full confidence: an invented fact, source, law, or…

I

  • Image generation : Image generation is AI creating pictures from a text description: you describe what you want, it draws it in…

L

  • LLM (large language model) : An LLM (large language model) is the engine behind tools like ChatGPT: a program trained on enormous amounts…

M

  • Multimodal : A multimodal AI understands and produces more than text: it can look at photos, read documents, listen, talk…

O

  • Open-weights model : An open-weights model is an AI model whose trained "brain" is published for anyone to download, run, and…
  • Opt-out : An opt-out is the setting that tells an AI provider to stop using your conversations to train its models. Most…

P

  • Personal data : Personal data is any information that relates to an identifiable person: a name, email, phone number, photo…
  • Prompt : A prompt is whatever you type (or say) to an AI: your question, request, or instructions. The clearer and more…
  • Prompt engineering : Prompt engineering is the craft of writing requests that get good answers from AI: giving context, examples…
  • Prompt injection : Prompt injection is an attack where malicious instructions are hidden inside content an AI reads (a web page…

R

  • RAG (retrieval-augmented generation) : RAG (retrieval-augmented generation) means the AI first looks things up in a trusted source (your documents, a…
  • Reasoning model : A reasoning model is an AI model that works through a problem step by step before answering, instead of…

S

  • Shadow AI : Shadow AI is when employees use AI tools at work without the company knowing or approving, pasting work data…
  • Synthetic data : Synthetic data is artificially generated data that mimics the patterns of real data without containing real…
  • System prompt : A system prompt is the hidden instruction sheet an AI receives before your conversation starts. It sets the…

T

  • Temperature : Temperature is a setting that controls how predictable or adventurous an AI's wording is. Low temperature…
  • Temporary chat : A temporary chat is a conversation mode where the AI doesn't save the chat to your history or use it for…
  • Token : A token is the small chunk of text an AI actually reads and writes: roughly three-quarters of a word in…
  • Training data : Training data is the huge collection of text, images and code an AI model learns from. With some chatbots…
  • Two-factor authentication : Two-factor authentication (2FA) protects an account with a second proof beyond the password: a code from an…

U

  • Usage limits : Usage limits are the caps an AI service puts on how much you can use it: messages per hours, file uploads…

V

  • Voice mode : Voice mode lets you talk to an AI assistant out loud and hear it answer, like a phone call. Available in…