Show a desktop notification when the AI TA finishes replying
Notify me when classmates post messages in the forum
Play an alert sound whenever there is a new notification
Use AI, but do not let AI use you
AI has already permeated every corner of learning, work and daily life. AI literacy is not about learning more AI tool functions, but about learning to keep your own judgement and creativity, and not being led by AI in the process of using it.
This page is written by Uedu An integration of six official AI literacy documents from international and local sources (UNESCO, OECD, EU AI Act, AAAI, IEEE, Taiwan Ministry of Education),Mainly written for university students, teachers and those concerned about AI education, and it is also suitable for readers at senior secondary level and above, or anyone who wants to understand what AI literacy is. Want to view the academic version →
47-item self-assessment, about 10–15 minutes. After completion, you will receive a personal AI literacy spectrum — clearly showing where you stand on each of the six capability dimensions, which areas are already solid, and which still need strengthening. Your quiz results will also be combined with your learning activities across Uedu courses, forming your own AI literacy trajectory.
Before entering AI literacy, first clarify what 'literacy' means
Have you heard of “media literacy”? In recent years, the news, schools and parents have all been talking about it — teaching people to identify fake news, spot bias and verify sources.
"Media literacy" is a type of literacy; people are just more familiar with that name.
You can finish reading one message.
You will ask who wrote it, what the stance is, and whether it should be checked before sharing.
You will use AI, and:
It is the same capability as media literacy —just the AI version.
It is not because AI is so frightening, but because three things happen every day
You turn on AI to help with your homework. It writes fluently and sounds very convincing. Only after submitting it do you realise the cited paper does not exist at all.
You look up a historical event, and AI gives you a time, place and person that sound perfectly reasonable. How do you know it is true? Or is it “confidently making things up”?
AI designed a poster for you, wrote a piece of code, and drew a picture. It looks good. But is it your work? Can you say, 'I made this'?
There is no standard answer to these three questions. Butpeople with AI literacy can handle them calmly.
Over the past three years, from top universities in the United States to the United Nations, the world has kept moving —
The Board of Trustees of Purdue University approved a policy on 12 December 2025 — starting with first-year students admitted in autumn 2026, all undergraduate students must have “AI Working Competency” before graduation. The main campus (West Lafayette + Indianapolis) has 57,000+ students. This is the first university in the United States to make AI literacy a graduation requirement.
Principal Mung Chiang stated clearly at the university senate: "Many companies have stopped hiring and are even carrying out large-scale layoffs. This forces universities to think—which jobs will be replaced by AI? Which jobs will be replaced by people who know how to use AI?"
This is not educational reform; it is a survival issue for employment.
The EU has passed the AI Act (Regulation 2024/1689). The law explicitly requires organisations using AI systems to ensure that their staff have AI literacy; breaches of the provisions carry penalties.
The United Nations Educational, Scientific and Cultural Organization (UNESCO) has released the AI Competency Framework for Students, clearly listing what students need to know in the AI era.
Leading universities have successively formulated guidelines for the use of generative AI, introduced general education courses on AI literacy, and established AI ethics committees. Stanford established the HAI (Human-Centred AI) research centre as early as 2019 and has invested substantial resources.
The AI academic organisation AAAI’s Long & Magerko proposed the AI Literacy framework at CHI 2020; to date it has been cited by more than 800 academic papers and is a key reference for university course design worldwide.
Cornell University launched an online AI critical thinking module, and 7,000+ students have completed it. The focus is not on teaching how to use AI, but on cultivating the ability to 'think in the age of AI' — giving students a shared language.
DeVry University has publicly committed that, by the end of 2026, every course will embed AI literacy training. From nursing and accounting to engineering, teaching content across the whole university must be updated.
Not only overseas — Taiwan itself has a series of local initiatives as well —
ChatGPT appeared at the end of 2022, and GPT-4 was released in 2023 — the pace of capability leaps caught the world unprepared. Since then, governments, international organisations and universities have spent three years rushing to fill the gap.
World Economic Forum (WEF) 2023 Future of Jobs Report conclusion: "AI and big data" are the top priority skill training focus for the next five years.
Not using AI will not get you phased out —
what will is not being able to “use AI” (that is, lacking AI literacy).
From understanding AI to rejecting AI, six capabilities form complete AI literacy. None of these capabilities are abstract; they are all judgements you will use in everyday life.
You know what AI is doing, and what it is not doing. You do not treat AI as magic, nor as a search engine for looking things up.
Can I state the key difference between ChatGPT and Google Search?
Know that AI can be biased and can make mistakes. But when you use something produced by AI, the responsibility is not AI's, it is yours. There is another often overlooked aspect — the electricity used to train a large AI model is roughly equivalent to a small town's annual electricity consumption. 'Should AI be used' is not just an academic question, but also an environmental responsibility.
If AI wrote something incorrect in my assignment and the teacher caught it — is that AI's responsibility or mine?
You have the right to know where the data you enter into AI goes and how it is used. Not everything should be handed over to AI.
Do I know where my conversations with AI are stored and what they are used for?
AI often gets things wrong, but sounds very confident. Being able to tell when to trust it and when to verify is a core skill.
When I see an AI answer, should I first ask “Is this correct?” before using it?
It is not about asking AI to write for you, but about helping AI think more deeply and helping you learn more solidly. Treat AI as scaffolding, not a substitute.
I use AI not to "understand" better, but to "skip learning"?
Can use AI to create things that were not previously possible — while also sticking to one's own judgement when the advice from AI is wrong.
Am I willing not to accept AI advice when it is wrong, rather than taking it at face value?
Are six capabilities too abstract? Try applying them to the 6 things you do most often — the difference in the same activity with and without AI literacy.
Literate use: Use AI to "test your own ideas" — let AI challenge you and identify weaknesses in your argument.
Lack of literacy: Letting AI think for you and accepting everything it gives.
Literate use: Use AI to explore multiple solutions, then judge for yourself which suits your context.
Lack of literacy: Giving the problem to AI, asking for one answer, and following it.
Literate use: Use AI to help you understand difficult passages — ask it to explain, give examples and probe further.
Lack of literacy: Asking AI to write the summary instead of reading it yourself.
Literate use: Keep your own voice; use AI for grammar and structural suggestions.
Lack of literacy: Submitting AI-written work as your own.
Literate use: Use AI to find a direction and get started — then verify the original sources yourself.
Lack of literacy: Using what AI gives you as if it were "research you found" — even fake citations and you would not know.
Literate use: Treat AI as a co-creator — you decide the topic, style and trade-offs.
Lack of literacy: Directly taking AI-generated work as your own without disclosure or responsibility.
The key issue is not whether to use AI, but how to use it — with the same tool, the results produced by someone with digital literacy and someone without can differ greatly.
AI literacy is not something you either have or do not have; it is a journey. It grows gradually through three stages.
You know what AI is, what it can do, and what it cannot do. Like just getting a new tool, you start by getting to know it.
You can use AI to achieve your goals in study, work and everyday life. It is like using a tool naturally, with improved efficiency.
You can use AI to create things no one has made before. It is like using tools to produce new works, new workflows and new possibilities.
No one starts at stage 3. For each capability area, you will be at different stages. For example, you may already be at stage 3 in "AI Awareness", but still at stage 1 in "Data Protection"—that is normal.
These are the phrases you hear most often about AI literacy — but they are all not quite right.
From Purdue University’s 2025 official AI usage guidelines — simple enough to memorise.
State which AI model and which version you used. For example: “ChatGPT-4o”, “Claude Sonnet 4.5”, “Gemini 2.0”.
Keep a record of your conversations with AI, or screenshots. If something goes wrong, you can reproduce your thought process and prove that you did not copy directly.
Explain how you verified the information given by AI, what errors you found (AI hallucinations) and corrected, and what further judgements and content you added yourself.
Answer each question in your head first. Yes, unsure, no — all are normal; see where you are now.
Can I explain clearly how AI differs from search engines and from “information on the internet”?
When I see an AI answer, I first think, "Could there be bias here?"
Do I know, before pasting data into AI, whether I should paste it and which kinds of data should not be pasted?
When I look at AI answers, I actively verify one or two things before believing them.
I use AI not just to save time, but to help me learn more thoroughly.
When I think AI suggestions are wrong, I say “no” and stick to my own judgement.
It is not another ChatGPT — the way we designed it is to “use AI to train your AI literacy”.
Most AI tools on the market are designed to make you use them more — the more you use them, the better; the faster the answers, the better. Uedu is different. We want you, after using Uedu, to think more critically, be more willing to challenge AI, and know more clearly when you should stop and think for yourself. Here is how we do it.
When you ask Uedu's AI a question, it often does not give you the answer directly — instead it asks back, 'What do you think?' and 'What have you tried?' so that you work it out yourself. What this trains is your brain, not the AI's.
Facts, citations and data mentioned by AI on Uedu will, as far as possible, be attributed to their sources (which book, which publication, which database).You can click to verify them without having to believe that "what AI says is correct".
After you enter Uedu, the platform will ask you to complete a few 'learning traits' mini quizzes (for example, your personality, and how you learn best).AI will then interact with you in a way that suits you—rather than giving everyone the same standard answer.
After your conversation with Uedu's AI ends, the platform will analyse the level of your thinking today — whether you are only 'remembering answers', or have moved on to 'analysis' and 'creating on your own'.So you can see where you have improved, rather than being led by the AI.
On every class on Uedu, the teacher clearly sets "how AI should be used in this class" — whether it is prohibited, allowed with conditions, or encouraged (the Purdue three-mode model).No need to guess; you will know exactly where the red line is.
AIDA is Uedu's learning companion—it remembers your learning history and knows where you get stuck, supporting you in discussions across courses. It is not a tool to do questions for you, but a companion that helps you think independently more with use.
Key difference: other AI tools do everything they can toget you to use more; Uedu does everything it can tohelp you think better—even if one day you leave Uedu, that capability stays with you.
Special note for teachers, policymakers and parents — this is the next question arising from AI literacy.
For final exams, written reports, and course assignments — these ‘final deliverables’ — students can now use AI to write them all. Reports may look fluent, well structured, and well referenced — but you cannot tell which ones are genuinely student work and which are written by AI. Traditional assessment is losing its most important function: identifying who has really learned.
Formative assessment looks not at the final 'output', but at the student's 'thinking process' — where they get stuck, whether they ask follow-up questions, whether they can challenge AI's answer, and whether they can correct it when AI gets it wrong. Process evidence does not lie.
Note: formative assessment supplements rather than replaces summative assessment — traditional final examinations still have value, and both coexist.
Uedu's 'knowledge trail' automatically records every step of a Student's conversation with AI, every follow-up question and every turning point — allowing your class to naturally accumulate the evidence needed for formative assessment.
Academic positioning: Uedu’s knowledge trail provides the process evidence needed for formative assessment, but is not the same as formative assessment itself. A complete formative assessment requires learning objectives, assessment criteria, and teaching adjustments by the Instructor.
You do not need to be an AI expert — but there are some things you can do right away.
All international frameworks, university cases and Taiwan initiatives integrated on this page are linked below as official links that can be opened directly.
This page contains no vague generalities — all legislation, policies, university cases and statistical data come from the public sources listed below. If you are doing research, writing a report or making policy recommendations, please refer to the original links.
Cite this page: This page’s content is compiled by Uedu from the international and local public sources above and may be freely cited; please attribute it as “Uedu AI Literacy Plain-language Guide, uedu.tw/ai-literacy”.
Uedu is a platform that lets you learn with AI without being replaced by AI. Every class, every conversation with AI, is an opportunity to practise AI literacy.
Get started with Uedu Learn about Uedu's research framework