Free Course Guide · Last updated October 20, 2026
A free AI course for work you can start this week, no account needed
Night school logic: learn in the gaps of a busy week, then use it the next morning.
This guide is for people who already have a job and want a free AI course for work that teaches them to use AI on emails, reports, meeting notes, spreadsheets and documents, not theory. It uses Anthropic's Claude as the assistant you practice with, walks you through a first week you can fit around full-time work, and explains how to practice safely and what a free course will not give you.
Yes, there is a free AI course for work. Start with the free Meet Claude lesson (6 minutes, no account needed), then take The Three Ways to Use Claude and The Iteration Loop. Each lesson is short enough for a lunch break. After each one, apply it to a real task such as an email, a meeting summary or a report, using non-confidential material, and check every result before anyone else sees it.
Published
Free lesson · no account neededMeet ClaudeWhy Claude?6 minStart the free lessonKey Takeaways
- Start with the free Meet Claude lesson, which needs no account. Then take The Three Ways to Use Claude and The Iteration Loop, in that order.
- A course is useful for work when it makes you practice on your own tasks the same day, not when it explains how AI models are built.
- Do not paste confidential material into a personal account. Read your employer's AI policy first and practice with public or invented material.
- Free lessons teach habits: how to ask, how to improve a draft, how to check. They do not teach your company's tools, data or approval rules.
- You stay responsible for everything you send. Verify names, numbers and claims before AI-assisted work reaches a colleague or client.
Useful free AI training for work versus a waste of time
| What to look at | Useful for work | Likely a waste of time |
|---|---|---|
| Starting point | You can read the first lesson right away, without an account, and decide if the style suits you. | You must give personal details and sit through a sales pitch before you see any content. |
| Lesson content | Concrete tasks: replying to an email, summarizing notes, improving a draft, analyzing a document. | History of AI, definitions of model types and predictions about the future, with nothing to try. |
| Practice | Each lesson ends with something you can do on your own work the same day. | You watch or read and are left to work out how it applies to your job. |
| Length and pace | Short lessons you can finish in a lunch break, at your own pace, in any order of days. | Long sessions that demand a fixed schedule you cannot keep alongside a job. |
| Honesty about limits | Teaches you to check output, spot invented facts and decide what not to delegate. | Treats the assistant as always right, or treats it as magic. |
| Data safety | Reminds you to keep confidential material out of personal accounts and to follow your employer's rules. | Says nothing about what you should and should not paste into a chat. |
Copy and Paste
Prompts You Can Use Today
Reply to a customer or colleague email
I am a [job title] at a [type of company]. Below is an email I received, with all names removed. Write a reply in a calm, polite tone, under 150 words, that answers each question in the order they were asked and proposes one clear next step. Afterwards, list any question you could not answer from the text, so I can check it myself. Email: [paste email here]
Replace the job title and company type with yours. Remove real names, account numbers and anything private before pasting.
Turn rough meeting notes into decisions and tasks
Here are my rough notes from a team meeting, with people shown only as initials. Turn them into a summary with three parts: decisions made, tasks with an owner and a due date, and open questions. If a task has no owner or no date in my notes, write 'missing' instead of guessing. Keep it short enough to send in one message. Notes: [paste notes here]
The instruction to write 'missing' instead of guessing is the important part. It stops the assistant from inventing owners and dates.
Get help with a spreadsheet using fake data first
I work in Excel (or Google Sheets) with a table that has these columns: date, product, units sold, region. Explain step by step how to build a summary of total units by region and by month using a pivot table. Then give me one formula I can use to check that the summary total matches the original data. Use a few invented example rows so I can test the steps before touching my real file.
Swap in your own column names. Never paste a real customer or payroll table into a personal account.
Summarize a public report and force it to show its sources
Below is a public report I need to understand before a meeting. Summarize it for a manager who will only skim it, in five bullet points. After each bullet, quote the exact sentence from the report that supports it. Then list three questions a careful reader might ask that the report does not answer. Report text: [paste a public document here]
Use a document that is already public, such as an industry report or a government publication. Check the quoted sentences against the original.
Improve your own draft without letting it rewrite everything
Here is my draft of a weekly status report for my manager. Do not rewrite it. First tell me the three weakest parts and explain why each one is weak. Then suggest one specific change for each. I will decide which changes to keep, and then I will ask you to revise only those parts. Draft: [paste your draft here, with private details removed]
This practices the habit taught in The Iteration Loop: review, decide, then refine in small steps.
Free to start
Start with a free lesson, no account needed
Take these three in order. Each one is short, and each leads to something you can try on your own work the same week.
What to look for
What makes free AI training for work worth your time
Most people searching for a free AI course for work are not short of information. They are short of time and of a clear first step. These six qualities separate a course that changes how you work from one that only fills an evening.
Short lessons you can actually finish
A lesson that fits in a lunch break or a commute gets done. A lesson that needs a free afternoon waits for months.
Your real tasks, not toy examples
Emails, status reports, meeting notes, spreadsheets and contracts are the material of most jobs. A good course uses material like that.
Practice the same day
A skill sticks when you use it within hours of learning it. Look for lessons that give you something to try on today's work.
Honesty about mistakes
AI assistants can sound sure and still be wrong. A useful course teaches you how to check, not only how to ask.
No technical background assumed
An AI course for beginners at work should start from writing a clear request in plain language, the same skill you use to brief a colleague.
A self-paced structure
You decide when to learn. A self-paced AI course lets you stop on a busy week and resume the next one without losing your place.
The starter path
Where to begin: the free starter lessons, in order
These lessons are free and readable without an account. Take them in this order, and try something on your own work after each one.
1. Meet Claude (course: Why Claude?)
A 6-minute lesson that introduces the assistant you will practice with. It is the first lesson in the Why Claude? course, and it needs no account.
2. The Three Ways to Use Claude (course: The Three Modes)
A 5-minute lesson on the different ways you can work with the assistant, so you pick the right one for a task instead of using one for everything.
3. The Iteration Loop (course: Master the Essentials)
A 6-minute lesson on improving a result step by step. This is the habit that turns a mediocre first answer into something you can use at work.
Then: Document Analysis
A 7-minute lesson in Master the Essentials. It is a natural next step when your daily work involves long reports, contracts or attachments.
Then: Context Is Everything
A 5-minute lesson in the Strategic Prompting course, about giving the assistant the background it needs so you stop getting generic answers.
Your weekly rhythm
A weekly rhythm that fits around a full-time job
You do not need a free week to learn AI for your job. You need a few small, repeatable sessions. Think in terms of an hour or two per week, split into short blocks: one on a lunch break, one on an evening, and one on a weekend morning if you want a longer practice run.
Each block has a different purpose. The lunch break is for reading one short lesson. The evening is for trying what you read on a real task. The weekend block is for reviewing what worked, saving the prompts you want to reuse, and choosing next week's task. This keeps the learning tied to your work instead of floating free of it.
Many people stall because they treat learning as one big event. A self-paced AI course works the other way around. If a week goes badly, you lose nothing. Open the next lesson when you have a spare half hour, not when you have a perfect plan.
A useful rule is to never read a second lesson before you have used the first one on something real. Two lessons read and zero tasks tried is the most common way people feel they are learning while their work stays the same.
First week, part one
Your first week, step by step: days one to three
Step 1, day one: take the free Meet Claude lesson from the Why Claude? course. The Meet Claude lesson is 6 minutes and needs no account. When you finish, write down one task from your job that you do often and that does not involve secrets: a weekly update, a polite reminder, a summary of a public document.
Step 2, day one or two: define what a good result looks like before you open any tool. Write two or three lines: who will read it, what tone it needs, how long it should be. This small step is the difference between asking for help and hoping for luck, and it is the base of every prompt you will write.
Step 3, day three: take The Three Ways to Use Claude, the first lesson of the course The Three Modes. The Three Ways to Use Claude lesson is 5 minutes. Afterwards, decide which way suits your task. Many working tasks only need a simple chat; others benefit from a different approach. Knowing the options means you choose on purpose instead of by habit.
If you do not have an account for the assistant yet, this is the point at which you may want one. Use a personal account for practice, with material that is safe to share. The next section explains how to decide what is safe.
First week, part two
Your first week, step by step: days four to seven
Step 4, day four: do the task for real, using a safe version of your material. Paste in your own rough draft of an email, an invented set of meeting notes, or a public document. Use the first prompt example on this page as a model. Do not try to get a perfect answer on the first attempt.
Step 5, day five: take The Iteration Loop, the first lesson of Master the Essentials, which is 6 minutes. Then go back to your result from day four and improve it in rounds. Ask for the three weakest parts, choose what to change, ask for a revision, and compare. Two or three rounds usually matter more than a longer first prompt.
Step 6, day six: check the result as if a colleague had written it. Verify names, dates and numbers against your original. Read the tone aloud. Note how long the task took with help compared to without it. If it saved time, save the prompt in a note for next time. If it did not, write down why.
Step 7, day seven: review and plan. Look at what you learned: which kinds of tasks the assistant helped with, which it did not, and where you had to correct it. Choose the next task for the following week, and pick your next lesson from the path below. That is a complete first week, built from a few short sessions.
What to try first
Real tasks to practice on while you learn AI for your job
Choose the task that annoys you most, as long as it can be done with material that is safe to share. These five cover most office jobs in any country.
Emails and messages
Draft replies, shorten a long message, or change the tone from blunt to polite. You send the final version, so you keep control of what is said.
Meeting notes
Turn rough notes into decisions, tasks and open questions. Ask the assistant to mark missing owners and dates rather than guess them.
Spreadsheets
Ask for help with formulas, pivot tables and cleaning messy columns in Excel or Google Sheets. Test on invented rows first, and verify totals yourself.
Reports and documents
Summarize a long public document, outline a report, or tighten a draft so it reads clearly. Ask for supporting quotes so you can verify.
Research and preparation
Prepare for a meeting by asking what questions a supplier, client or manager is likely to raise. Treat the answers as prompts for your own thinking, not as facts.
Practice safely
How to practice on real work without exposing confidential data
Using AI at work with company documents is a question for your employer, not for a course. Many organizations have a written AI policy, an approved tool, or a rule that bans certain tools entirely. Before you paste anything from your job, read that policy or ask your manager, your IT team or whoever handles data protection. If no policy exists yet, ask anyway, and assume the cautious answer until you hear otherwise.
A personal account is not the right place for confidential material. That includes client names and contracts, personal data about customers or employees, unpublished financial figures, passwords and internal strategy. Data protection rules, such as GDPR in Spain and the rest of the European Union and the national laws across Latin America, can apply to this kind of data. Your company may also have confidentiality duties of its own.
You can still practice on real work without those risks. Rewrite the task with invented names and numbers. Use public documents such as an industry report or a government publication. Paste your own writing, such as a draft with private details removed. Describe the shape of the problem in words instead of pasting the data, and ask for a method rather than a finished answer.
If your employer does provide an approved AI tool, move your practice there once you know what you want to do. The habits you build in a personal account, such as writing clear requests, iterating and checking, transfer directly. The difference is that the approved tool may be the only place where real company material is allowed.
Check before you send
Checking AI output before it reaches a colleague or client
An assistant can write a fluent paragraph that contains a wrong date, an invented source or a number that does not add up. Fluency is not accuracy. A reasonable rule is that anything with your name on it is yours, however it was drafted, so you read it as carefully as you would read a junior colleague's first draft.
Build a short checklist and use it every time. Compare every name, date and figure with your original. Recalculate any total yourself. Ask whether each claim could be verified, and if you cannot verify it, remove it or mark it as unconfirmed. Read the tone from the reader's point of view, and ask whether you would be comfortable if they knew an assistant helped draft it.
For spreadsheets, test a formula on a few rows where you already know the right answer before you apply it to a whole sheet. For summaries, ask the assistant to quote the sentences it relied on, then look those sentences up. For anything with legal, medical, financial or safety consequences, the assistant can help you prepare questions, but a qualified person must make the decision.
The lesson Claude's Strengths & Limits, 6 minutes in the Why Claude? course, is worth taking early for exactly this reason. Knowing where the tool tends to fail makes your checking faster and more targeted.
What comes next
After the starter path: choose the next free lesson by your job
Once you have finished the starter path and tried it on a real task, choose the next lesson by what your week looks like. If you work with long documents, take Document Analysis in Master the Essentials. The Document Analysis lesson is 7 minutes. The Research and Thinking lesson is 7 minutes and follows naturally. If your results feel generic, take Context Is Everything (5 minutes) in Strategic Prompting, and Claude Prompting Fundamentals for Everyday Work (5 minutes) from the same course.
If you want more of the foundations, the Why Claude? course has more free lessons: Your First Conversation, Context Windows & System Prompts, Claude's Strengths & Limits, From Chatting to Building, and Make Claude Sound Like You. The whole course is 6 lessons and 40 minutes in total, so it fits into a few lunch breaks. The Three Modes course is 5 lessons and 30 minutes, and Master the Essentials is 4 lessons and 27 minutes.
If your employer has already chosen another tool, there are first-week lessons for those as well. The Microsoft 365 Copilot course starts with a lesson called Clear your morning inbox before your next meeting. That lesson is 11 minutes. The Google Gemini course starts with Clear a Crowded Inbox With Gemini in Gmail. That lesson is 12 minutes. The ChatGPT course starts with Getting an answer you can actually use. That lesson is 14 minutes. The skills of asking clearly, iterating and checking carry across tools.
Not every lesson on the site is free, so check each lesson before you plan around it. The ones named in this guide are the free, anonymously readable ones.
Honest limits
What a free AI course will not give you
A few weeks of practice with free lessons will give you a working habit: you will know how to describe a task clearly, how to improve a first answer, and how to check the result. That is real, and it is enough to save time on routine writing and organizing. It will not make you an expert, and it will not replace a colleague's judgment on hard decisions.
A free course also cannot teach you your own company. It does not know your approved tools, your data rules, your templates, your clients or your manager's expectations. Those you must learn inside your workplace, ideally by talking to the people who own the policies, and by showing a small, safe example of what you did before asking for wider use.
It will not give personalised feedback on your specific drafts, and it will not guarantee results. Some tasks will improve a lot, some a little, and some not at all, such as work that depends on information the assistant cannot see. Tools also change, so lessons age; treat them as a starting point and keep practicing.
Finally, a course does not decide for you what should be automated. Some work is better done by hand, because the thinking is the point, or because the stakes are high. The best outcome of a first month is not doing everything with AI, but knowing which tasks are worth handing over.
Reading about AI and using it at work are different skills
The quickest way to find out if AI helps with your job is to try it on one real task. The free Nightschool AI lessons are hands-on from the first one, and you can begin at your own pace without creating an account.
Frequently Asked Questions
Is there a free AI course for work?
Yes. Nightschool AI offers free, short lessons that you can read without an account. For work, start with Meet Claude in the Why Claude? course, then take The Three Ways to Use Claude and The Iteration Loop. They are built around everyday tasks like emails, notes and documents. Not every lesson on the site is free, so check each lesson, but the starter path is.
What is the best free AI course for my job?
The best one is the one you finish and apply to your own tasks. Look for short lessons, practical examples and an honest account of the tool's limits. A good start is the Nightschool AI starter path: Meet Claude, The Three Ways to Use Claude and The Iteration Loop. If your employer uses Microsoft 365 Copilot, Gemini or ChatGPT, there are first-week lessons for those too.
Can I learn AI for work without paying?
Yes, you can start without paying. The free starter lessons are readable without an account, and they cover the habits that matter most: writing a clear request, improving a result in rounds and checking it. Some lessons and courses on the site are not free, so look at each one before you plan around it. The first week of practice does not depend on any of them.
How long does a free AI course take?
The starter lessons are short: Meet Claude is 6 minutes, The Three Ways to Use Claude is 5 minutes and The Iteration Loop is 6 minutes. The Why Claude? course is 6 lessons and 40 minutes in total, The Three Modes is 5 lessons and 30 minutes, and Master the Essentials is 4 lessons and 27 minutes. The real time goes into practice, usually an hour or two per week.
Do I need technical skills to take an AI course?
No. If you can write a clear email or explain a task to a colleague, you have the skill you need to begin. The lessons use plain language. One lesson in The Three Modes, Claude Code: Build With Words, is about building with words and is optional for most office jobs. You can ignore it until a task calls for it.
Can I use AI at work with company documents?
Only if your employer allows it, and usually only in a tool the company approves. Do not paste confidential documents, client data or personal information into a personal account. Read your company's AI policy or ask your manager or IT team first. While you learn, practice with public documents, invented examples or your own writing with private details removed.
What should I learn first to use AI at work?
Learn to describe a task clearly: who the reader is, what you need, the tone and the format. Then learn to improve a result in rounds instead of expecting a perfect first answer, and then learn to check the output. Those three habits are what the starter path covers, in the lessons Meet Claude, The Three Ways to Use Claude and The Iteration Loop.
Do I need to create an account to start?
No. The free lessons in the starter path, including Meet Claude, can be read without an account. You will want an account for the assistant itself when you begin practicing, and you can use a personal one for safe, non-confidential tasks. Create it when you are ready to try the first task, not before.
Is a free AI course enough to use AI at work?
It is enough to start and to save time on routine writing and organizing. It is not enough to cover your company's tools, data rules or specialised tasks. Treat the free lessons as the base, then learn your employer's approved tools and policies, and keep practicing on real tasks. Skill comes from repeated use, with checking each time.
What if my company blocks AI tools?
Respect the rule. You can still learn the skills away from work material: practice on public documents, invented examples or personal tasks such as planning a trip. If you think AI could help your team, prepare a small, safe example and ask your manager or IT team whether an approved option is possible.