AI Awareness for Everyone

A friendly, no-jargon introduction to AI for every employee, from the shop floor to the front office. Learn what AI is and is not, how chat tools work and why they can be wrong, how to use them for everyday tasks, and how to protect data, spot scams and follow your company's rules.

Course structure

• Module 1: What AI Is (and Is Not) (4 lessons). A plain-language starting point: what people mean by AI, the main kinds you will hear about, the AI you already use every day, and the myths worth dropping.
• Module 2: Generative AI: How It Works and Using It Well (5 lessons). How AI chat tools produce answers, why they can be confidently wrong, the everyday work tasks they help with, and simple habits for asking and checking.
• Module 3: Staying Safe: Data, Privacy and Scams (4 lessons). What you should never paste into public AI tools, why company-approved tools matter, Canadian privacy law in plain terms, and how to spot deepfakes and AI-assisted scams.
• Module 4: Using AI Responsibly (4 lessons). Bias and fairness, intellectual property, being honest about AI use, and a balanced look at how AI is changing jobs and skills.
• Module 5: Your Company, Your Habits (4 lessons). Understanding your company's AI policy, knowing when and how to ask for help, building AI confidence as a team, and a personal 30-day AI habit plan.

Interactive labs

• Fact, myth or it depends?: Sort 12 common statements about AI into Fact, Myth or It depends, then check your answers. It builds the habit of questioning big claims about AI.
• Is it safe to share?: Decide where each piece of workplace information can go: any AI tool, only a company-approved tool, or nowhere. It turns the data rules into a quick, everyday judgment.
• Your AI awareness check: Rate yourself or your team on seven areas of everyday AI awareness. Your radar chart shows strengths and gaps, with a next step for each area. Retake it after your 30-day plan.

How it works

• Work through 21 short lessons at your own pace. Each has key takeaways, an exercise for your own workplace and a quick knowledge check.
• Try the three interactive labs as you reach them in the course.
• Pass the 20-question final exam (70% or higher, unlimited retakes) to unlock your Chmltech certificate of completion.
• Free, with no sign-up. Your progress is saved in this browser.

You hear the word AI everywhere. On the news, in software ads, in the lunchroom. It can sound big and mysterious. It does not need to be.

In everyday terms, artificial intelligence (AI) is software that does tasks we usually think need human judgment. It can spot patterns, make a guess, sort things into groups, or write and talk in a way that sounds human.

Regular software follows exact rules a person wrote. If an invoice is over $5,000, send it to the manager. AI is different. Instead of following only written rules, it learns patterns from many examples. Then it uses those patterns to make a best guess about something new.

A simple way to picture it

Think about a new hire at a distributor's warehouse. On day one, they need a checklist for everything. After a few months, they can glance at a pallet and tell it is packed wrong. Nobody wrote a rule for every bad pallet. They learned from seeing lots of good and bad ones. AI learns in a similar way, from data instead of experience.

What AI is good at, and what it is not

The key idea: AI makes predictions and best guesses. It does not think the way you do. It has no goals, no feelings and no understanding of your customers. It is a tool. A very useful one, when a person stays in charge.

You do not need to know any maths or code to use AI well. You need three things: a clear idea of what it can and cannot do, good habits for checking its work, and care with information. That is what this course gives you.

People use a few AI words as if they mean the same thing. They do not. Knowing the difference helps you understand what a tool can really do.

Narrow AI

Almost all AI in use today is narrow AI. It does one kind of job well. A spam filter sorts email. A map app predicts traffic. A camera on a production line spots scratches. None of them can do the others' jobs. The idea of a general AI that can do anything a person can is a topic for debate and science fiction. It is not the tool on your computer.

Machine learning

Machine learning is the most common way to build AI. You show a computer many examples, and it learns the pattern. Show it thousands of emails marked spam or not spam, and it learns what spam tends to look like. Then it can judge new emails it has never seen.

Generative AI

Generative AI is a newer kind of machine learning. Instead of only sorting or predicting, it creates new content: text, images, audio or video. AI chat assistants are the best-known example. You type a request and they write a reply, a summary or a draft.

How they fit together

Think of nesting boxes. AI is the biggest box. Machine learning sits inside it. Generative AI sits inside machine learning. And today, all of them are narrow in the sense that they are built for certain kinds of tasks.

Imagine a clinic's intake desk. A scheduling system that predicts no-shows uses machine learning. A tool that drafts a reminder message uses generative AI. Both are narrow. Neither knows the patient the way the receptionist does.

When someone at work says "we are using AI," a good first question is: which kind, doing what job?

Many people say they have never used AI. Most of them use it every day. It is built quietly into tools you already know.

Everyday examples

• Email filters. Your inbox sorts spam and flags messages that look like phishing. It learned what junk looks like from huge numbers of examples.
• Maps and navigation. Your map app predicts traffic and suggests a faster route. It learns from patterns in travel data.
• Recommendations. Streaming services, online stores and news apps suggest what you might like next, based on what you and people like you have chosen.
• Voice assistants. Smart speakers and phone assistants turn your speech into text and try to work out what you want.
• Phone cameras and keyboards. Your camera sharpens photos and recognizes faces. Your keyboard predicts the next word.
• Banking. Your bank may text you when a purchase looks unusual. That alert often comes from a machine learning model.

At work, too

Many workplace tools now include AI features. A word processor may suggest how to rephrase a sentence. A video call app may offer live captions or a meeting summary. Accounting software may suggest how to code an expense. A warehouse system may forecast which items will run low.

Why this matters

Seeing AI you already use does two things. First, it makes AI less scary. You have been working alongside it for years. Second, it shows you its limits. Think about it:

• Your spam filter sometimes hides a real email.
• Your map sometimes sends you down a closed road.
• Your recommendations sometimes miss badly.

That is normal. AI makes best guesses, and guesses are sometimes wrong. The same is true of the newer AI chat tools. The habit you already have of checking your spam folder now and then is exactly the habit you need for all AI: trust it for speed, check it for accuracy.

AI attracts a lot of myths. Some make it sound magical. Some make it sound like a monster. Both get in the way of using it sensibly. Here are common ones, and what is closer to the truth.

Why myths are risky at work

Myths lead to real mistakes. Imagine a supervisor at a fabrication shop who believes AI is always right. They ask a chat tool about a safety standard and copy the answer into a toolbox talk without checking. If the answer is out of date or wrong, workers get bad safety guidance.

Now imagine an office manager who believes AI is useless. They spend two hours writing a routine letter that a chat tool could have drafted in seconds for them to edit. Both myths cost something.

"It depends" is often the honest answer

Many questions about AI do not have a simple yes or no. Is it safe to use AI for this? It depends on the tool, the data and the task. Will AI change my job? It depends on what your job involves. Getting comfortable with "it depends" is a sign of good judgment, not confusion.

A useful habit: when you hear a big claim about AI, good or bad, ask "how do we know that?" and "in which situations?"

AI chat assistants feel like talking to a person. Knowing roughly what happens behind the screen helps you use them well and spot their limits. No technical background needed.

A very well-read guesser

These tools are built on what is called a large language model. During training, the model read a huge amount of text: books, websites, articles and more. From all that reading, it learned how words usually follow each other.

When you type a question, the tool does not look up a single right answer. It predicts a likely next word, then the next, then the next, until it has a full reply. It is a bit like the word suggestions on your phone keyboard, but far more powerful.

What happens when you hit Enter

1. You write a prompt. Your question or request, plus any text you paste in.
2. The tool breaks it into pieces. These small chunks of text are called tokens.
3. It predicts a response. Word by word, based on patterns it learned.
4. It shows you the answer. Often in seconds.
5. You check and decide. This step is yours, and it is the most important one.

Things worth knowing

• It can give different answers to the same question. There is some built-in variety in its word choices.
• It only knows what it learned, plus what you give it. Some tools can also search the web or read files you upload, but they still make mistakes.
• It does not know your workplace. It has never seen your price list, your customers or your procedures unless you share them, and you should be careful what you share.
• It has a memory limit. In a very long chat, it can lose track of things said early on.

Here is a quick example. An engineering office assistant asks a chat tool, "Write a short note telling clients our office closes early on Friday." The tool predicts a polite, typical note. It does not know the actual closing time unless you say it. So you add: "We close at 2 p.m. this Friday, December 20." Better input, better output.

The most important thing to know about AI chat tools is this: they can be completely wrong while sounding completely sure. When an AI tool makes up facts, people call it a hallucination.

Why it happens

Remember, the tool predicts likely words. Its goal is a reply that sounds right. Most of the time, sounding right and being right line up. Sometimes they do not. When it lacks good information, it does not usually say "I don't know." It fills the gap with something that looks plausible.

Common types of confident mistakes

• Made-up facts. A date, a figure or a rule that sounds real but is not.
• Fake sources. A report title or web page that does not exist.
• Out-of-date information. Rules, prices or contacts that have changed.
• Wrong maths. Totals or percentages that look neat but do not add up.
• Wrong context. Rules from another country or industry presented as if they apply to you.

A worked example

Imagine a supervisor at a small food distributor in Ontario. She asks a chat tool, "How many paid sick days do employees get here?" The tool gives a clear, confident number. It might mix up rules from another province, a different year or a different country. Employment rules vary by province and change over time.

The right move: use the AI answer as a starting point at most, then check the official provincial source or ask HR. Never put a fact like that in a staff notice based only on an AI answer.

Where to be extra careful

Confidence is not evidence. The more it matters, the more you check.

Used well, AI chat tools can save you time on the routine parts of your day. Here are five everyday uses that suit almost any role, from the shop floor to the front office.

1. Drafting

Get a first version of an email, a notice, a job posting or a short procedure. You then edit it so it is accurate and sounds like you. A blank page becomes a rough draft in seconds.

2. Summarizing

Turn long text into key points. For example, a supervisor pastes a non-confidential, approved-to-share policy update and asks for a five-bullet summary for the team huddle. Always check the summary did not drop something important.

3. Brainstorming

Generate ideas when you are stuck: ways to reduce clutter at a workstation, themes for a safety week, questions to ask at a supplier meeting. You pick the ideas that fit.

4. Translating

Help with a quick translation or understanding a message in another language. This can help in diverse teams. For anything important, such as safety instructions or contracts, have a fluent person review it.

5. Checking tone

Ask whether a message sounds too harsh, too casual or unclear, and get a gentler or clearer version. This is useful before replying to an upset customer.

A worked example

Imagine Sam, a shipping lead at a distributor. He needs to tell customers that a holiday will delay deliveries by one day. He writes in an approved AI tool:

Write a short, friendly email to customers. Deliveries scheduled for Monday will arrive Tuesday because of the holiday. Apologize briefly and thank them. Keep it under 100 words.

He gets a draft, fixes one date, adds his name and sends it. It took two minutes instead of fifteen.

Good fit or poor fit?

The pattern is simple: AI does the first pass, you do the final pass.

A prompt is what you type into an AI tool. Better prompts get better answers. You do not need special tricks. A few simple habits make a big difference.

Five prompting habits

1. Say what you want. Start with the task: write, summarize, list, rewrite, explain.
2. Give context. Who is it for? What is the situation? "For new warehouse staff on their first day."
3. Set the format. Bullet points, a short email, a table, under 150 words.
4. Set the tone. Friendly, formal, plain language, reassuring.
5. Ask for changes. If the first answer is not right, say what to fix. "Shorter, and remove the jargon."

A worked example: before and after

Weak prompt:

Write about safety.

This gives a long, general essay nobody will read.

Better prompt:

Write a 5-bullet reminder for forklift operators at a small warehouse about checking their truck before each shift. Plain language, friendly tone, short sentences.

This gives something you can actually use, after you check it against your own procedures.

Treat it like a conversation

You rarely get the perfect answer on the first try. That is fine. Think of the AI as a fast helper who needs direction. Ask it to:

• "Make this shorter."
• "Give me three other versions."
• "Explain this like I'm new to the job."
• "What questions might a customer ask after reading this?"

What to leave out

Good prompts give helpful context, but not private details. You can say "a customer who received a late order" instead of using their name and account number. Describe the situation without exposing the people. We cover this in detail in the next module.

One more tip: if a task is big, break it into smaller asks. First ask for an outline, then ask for each section. Shorter requests are easier to check, and you stay in control of where the work is going.

Simple formula to remember: Task + Context + Format + Tone, then refine.

You have learned that AI can be confidently wrong. So checking its work is not optional. It is the skill that makes AI safe to use. The good news: checking is a simple habit, and you can build it quickly.

The four-step check

1. Read it all. Do not skim. Read the whole output as if a new coworker wrote it.
2. Verify the facts. Check names, dates, numbers, rules and sources against something you trust: your own records, an official website or a colleague who knows.
3. Fix and fit. Correct errors, remove anything you cannot confirm, and adjust it to your workplace and your voice.
4. Own it. Once you send or use it, it is your work. Would you be comfortable putting your name on it?

Quick questions to ask yourself

• Is anything here that I do not actually know is true?
• Do the numbers add up?
• Does this apply to Canada, my province and my industry?
• Is anything missing that the reader needs?
• Could this offend, confuse or mislead anyone?

A worked example

Imagine Priya, an office administrator at an engineering firm. She uses an approved tool to summarize a long, internal project update for the weekly bulletin. The summary reads well. But when she checks it against the original, she spots two problems. The summary says the project finishes in March; the original says May. And it leaves out that a site visit was cancelled.

She fixes both. Two minutes of checking saved an embarrassing correction later.

Match the check to the stakes

The rule to remember: AI can help you write it, but you are responsible for what it says.

The fastest way to get into trouble with AI is to paste the wrong thing into the wrong tool. Many free, public AI tools may store what you type. Depending on the tool and its settings, that text could be kept, reviewed or used to improve the service. Once it leaves your hands, you cannot pull it back.

The never-share list (for public tools)

• Personal information about people: names with addresses, phone numbers, birth dates, social insurance numbers, health details, pay or performance information.
• Customer and supplier data: customer lists, account numbers, order histories, contract terms, pricing agreements.
• Passwords and access details: logins, PINs, security codes, network details. Never, in any tool.
• Financial information: bank details, unreleased financial results, payroll files.
• Confidential business information: drawings, formulas, quotes, plans, internal reports, anything marked confidential.
• Anything you were told to keep private: legal matters, HR issues, investigations.

The "front page" test

Before you paste, ask: would I be comfortable if this text showed up somewhere public with our company's name on it? If not, do not paste it into a public tool.

How to still get help, safely

You can usually get the help you need without the sensitive parts. Remove or replace details:

A worked example

Imagine Marco, a receptionist at a physiotherapy clinic. A patient emails an upset complaint that includes their health history. Marco wants help writing a calm reply. He does not paste the email. Instead, he types: "Help me write a calm, caring reply to a patient upset about a long wait time. Under 120 words." He gets a good draft and adds the personal touches himself.

When in doubt, leave it out, and ask your supervisor.

Not all AI tools are the same when it comes to your data. Many companies choose specific AI tools for staff to use. These are often called company-approved or sanctioned tools.

Why approved tools are different

When a company sets up an approved tool, it can usually:

• Sign a business agreement with the provider about how data is handled.
• Choose settings so your inputs are not used to train the provider's models.
• Control who has access and keep records of use.
• Connect it safely to company systems.

That is why some information that must never go into a public tool may be fine in an approved one. But approved does not mean anything goes. Your company's policy still decides what is allowed.

Three levels to keep in mind

Watch out for "shadow AI"

Shadow AI means using AI tools at work that the company has not approved. It often starts with good intentions. Someone finds a handy browser add-on or a free app and starts pasting work documents into it. The problem is that nobody knows where that data goes.

Imagine a 30-person engineering office. One drafter installs a free AI note-taker that joins video calls automatically. It records client meetings and sends summaries to an outside server. Nobody approved it, and clients never agreed. That is a real risk, created in minutes.

What to do

• Find out which AI tools your company approves.
• Use your work account, not a personal one, for work tasks.
• If you find a useful new tool, ask before using it with work data.
• If there is no approved tool yet, stick to public information only.

In Canada, privacy laws set rules for how businesses handle personal information. You do not need to be a lawyer, but every employee should know the basics, especially now that AI tools make it easy to copy and paste data.

What PIPEDA is

PIPEDA stands for the Personal Information Protection and Electronic Documents Act. It is the federal law that covers how many private-sector businesses collect, use and share personal information in the course of commercial activity. Some provinces, including British Columbia, Alberta and Quebec, have their own similar private-sector privacy laws that apply instead in many situations. Your company will know which apply.

What counts as personal information

Information about an identifiable person. For example:

• Name together with contact details
• Age, ID numbers, health information
• Employee records, pay and performance notes
• Customer purchase history linked to a person

The main ideas, in plain words

How this connects to AI

Pasting a customer's details into a public AI tool can mean sharing personal information with an outside company for a purpose the customer never agreed to. That can be a privacy problem, even if you meant well. Businesses may also have to report certain privacy breaches that create a real risk of significant harm.

Imagine a clinic receptionist who pastes a list of patient names and appointment notes into a free chat tool to "tidy the formatting." It took ten seconds and felt harmless. It may now be a privacy incident the clinic must deal with.

This lesson is general awareness, not legal advice. Your company's privacy officer or policy has the details. If you think personal information has been shared by mistake, report it right away. Quick reporting helps limit the harm.

Criminals use AI too. It helps them write convincing emails, clone voices and fake videos. The good news is that simple habits still stop most scams.

What has changed

• Better phishing emails. Scam emails used to be full of spelling mistakes. AI can now write clean, professional messages that sound like a real supplier or manager.
• Voice cloning. With a short sample of someone's voice, scammers can create fake phone calls or voicemails that sound like your boss.
• Deepfake video and images. AI can make realistic videos or photos of people saying or doing things they never did.
• Fake chat and websites. Lookalike websites and chat bots can trick you into entering passwords or payment details.

Warning signs to watch for

Since scams look more polished, focus less on spelling and more on behaviour:

1. Urgency. "This must be done in the next hour."
2. Secrecy. "Don't tell anyone, this is confidential."
3. Unusual requests. Buying gift cards, changing bank details, sending files to a new address.
4. A new channel. A manager who never texts suddenly texts you.
5. Pressure to skip steps. "Don't worry about the usual approval."

A worked example

Imagine Linh, who handles accounts payable at a mid-sized distributor. She gets a voicemail that sounds exactly like the owner: "I'm at the airport. Pay the new supplier invoice I'm emailing you today. It's urgent." An email arrives with new bank details.

Linh pauses. The request is urgent, unusual and skips normal approval. She calls the owner back on the number she already has on file, not the one in the email. The owner knows nothing about it. Linh reports it to IT. A scam stopped by one phone call.

Your simple defence

Some teams agree on a simple code word for unusual money or data requests made by phone. It is low-tech and effective.

AI learns from data. Data comes from people and the past. If the past was unfair, AI can learn that unfairness and repeat it, sometimes at scale and in ways that are hard to see.

How bias gets in

• Unbalanced data. If most examples come from one group, the AI may work poorly for others. A voice tool trained mostly on certain accents may struggle with others.
• Past decisions. If an AI learns from old hiring decisions that favoured certain people, it may keep favouring them.
• Stereotypes in text. Generative AI read huge amounts of text that include stereotypes. It can repeat them, for example by assuming a nurse is a woman or an engineer is a man.
• Missing context. The AI does not know your community, your customers or your team.

Where it matters most

Bias matters most when AI output affects people. Be especially careful with:

• Hiring, screening resumes or writing job ads
• Performance reviews and scheduling
• Customer service decisions, credit or pricing
• Images or content that represent people

A worked example

Imagine Dana, an office manager at a fabrication shop, asks an AI tool to draft a job ad for a welder. The draft says "he will be responsible for" and asks for "a young, energetic team player." Dana spots the problem. The language could discourage qualified applicants and may run against human rights rules. She rewrites it to use "you" and focus on the actual skills: welding certification, reading drawings, working safely.

Simple fairness checks

AI should never be the only decision-maker about a person's job, pay or opportunities. People decide; AI can assist.

Intellectual property (IP) means creations of the mind that the law can protect: written work, designs, logos, inventions, software and trade secrets. AI raises new questions about IP, and many of them are still being worked out by courts and lawmakers.

Two directions to think about

1. Protecting your company's IP going in. Drawings, formulas, source code, pricing models and proposals are valuable. If you paste them into a public AI tool, you may lose control of them. This is the same rule from the data lessons: keep confidential business information out of unapproved tools.

2. Respecting others' IP coming out. AI output can sometimes closely resemble existing work. An AI image might look like a known artist's style or a brand's logo. AI text might repeat phrases from a published source.

Practical rules

• Do not ask AI to copy a specific person's or company's work, logo or branded content.
• Do not present AI-generated images or text as original design work for a client without checking your company's rules.
• Check that important AI-written content is not copied word for word from somewhere else.
• Treat AI-generated code, designs and marketing material carefully, and follow your company's review process.
• Remember that who owns AI-generated content can be unclear, and tool terms of use differ.

A worked example

Imagine Ravi, who runs social media for a small building supply company. He asks an AI image tool for "a hardware store mascot like the one from a famous chain." The result looks very close to that chain's well-known character. Ravi realizes that using it could cause a trademark problem. He asks instead for an original design based on the company's own colours and a friendly beaver character. Then he shares it with his manager for approval.

This is general awareness, not legal advice. When IP questions come up, ask your manager.

Using AI at work is not cheating. It is a tool, like spell check or a calculator. But honesty about how you use it builds trust with your team, your customers and your company.

Why honesty matters

• Trust. People feel misled if they learn later that something personal or expert was mostly machine-written.
• Quality. If your manager knows AI helped, they know what extra checks may be needed.
• Responsibility. Being open keeps it clear that you, not the tool, stand behind the work.
• Policy. Some companies, clients and contracts require you to disclose AI use.

When to say you used AI

What honest AI use looks like

• Never pass off AI-generated facts as things you personally checked when you did not.
• Do not use AI to write fake reviews, fake testimonials or messages pretending to be someone else.
• Do not use AI to impersonate a coworker or customer.
• Be open with your manager about how you use AI in your work.

A worked example

Imagine Tom, a supervisor at a packaging plant. He uses an approved AI tool to draft a summary of a month's downtime notes for the operations meeting. He checks the summary against the logs and corrects two items. When he presents it, he says, "I used our AI tool for the first draft and checked it against the logs." Nobody minds. The team now knows the numbers were verified, and two coworkers ask him to show them how.

A good rule: if you would feel uncomfortable if people knew AI wrote it, either change how you used AI, or tell them.

Many people wonder: will AI take my job? It is a fair question. The honest answer is that nobody knows exactly how things will play out. What we can see is that AI is changing tasks within jobs, and that changes which skills matter.

Tasks, not whole jobs

Most jobs are a bundle of tasks. AI tends to help with some of those tasks, usually the routine ones: drafting, sorting, searching, summarizing and data entry. Other tasks stay firmly human: building relationships, handling the unexpected, hands-on work, leadership and judgment.

Imagine a customer service representative at a distributor. AI might draft replies to common questions and summarize long email threads. The representative now spends more time solving tricky problems and calming upset customers, the parts of the job that need a person.

A balanced view

Skills that grow in value

• Judgment: knowing when an answer is wrong or does not fit.
• Communication: explaining clearly, listening and building trust.
• Problem-solving: handling situations nobody planned for.
• Know-how about your work: the deep knowledge of your process, products and customers.
• Using AI well: asking good questions, checking outputs and handling data safely.
• Learning: being willing to try, adjust and keep growing.

What you can do

You do not need to become an AI expert. Start small. Learn the tools your company approves. Notice which of your tasks AI helps with and which it does not. Share tips with coworkers. Talk to your supervisor about training. People who understand both their work and how to use AI sensibly are valuable to any business.

Keep your own skills sharp, too. Use AI to help, not to stop thinking. Your experience is what makes AI output useful.

Many organizations now have an AI policy or acceptable-use guideline. It tells staff how AI can and cannot be used at work. If your company has one, it is your most important guide. If it does not yet, the habits in this course are a safe starting point until it does.

What an AI policy usually covers

How to read it

1. Find it. It may be in the employee handbook, on the intranet or with HR.
2. Look for the approved tool list. This tells you where to start.
3. Check the data rules. Match them to the kind of information you handle.
4. Note the contact. Who to ask with questions.
5. Check for updates. AI policies change often as tools change.

A worked example

Imagine a 60-person home care agency. Its one-page AI guideline says: staff may use the company's approved AI assistant for drafting and summarizing; client names and health details must never go into any AI tool; all client-facing content must be reviewed by a supervisor; questions go to the operations manager. A care coordinator reads it in five minutes and now knows exactly what she can do.

If there is no policy yet

Use these safe defaults:

• Use only public, non-sensitive information in AI tools.
• Check everything before you use it.
• Tell your supervisor how you are using AI.
• Suggest that the company create a simple policy. You could be helping everyone.

A policy is not there to stop you. It is there so you can use AI confidently and safely.

Nobody expects you to know everything about AI. Tools change fast, and questions are normal. Knowing when to ask and how to report problems is one of the most valuable AI skills you can have.

When to stop and ask

• You are not sure if information is OK to put into a tool.
• You found a new AI tool or feature and want to use it for work.
• An AI output will be used for a decision about a person, money or safety.
• Something in the AI output seems wrong, biased or strange.
• A customer asks whether you use AI, and you are not sure what to say.

When to report right away

• You or someone else pasted sensitive information into an unapproved tool.
• You received a suspicious message, call or video that might be a scam or deepfake.
• AI-generated content with errors was sent to a customer or posted publicly.
• A tool behaves unexpectedly, such as sharing data with others.

Who to go to

Mistakes happen: report them fast

Imagine Jordan, a junior estimator at a fabrication shop. He pastes a customer quote into a free AI tool to fix the formatting, then remembers the policy. He feels embarrassed and is tempted to say nothing. Instead, he tells his supervisor that afternoon. The supervisor works with the owner to check the tool's data settings and decide whether the customer needs to be told. Because Jordan spoke up quickly, the company could act.

Good workplaces treat honest reporting as a positive. Hiding a mistake is almost always worse than the mistake itself.

How to report clearly

1. What happened, and when.
2. Which tool or message was involved.
3. What information may be affected.
4. What you have done so far.

AI skills grow fastest when people learn together. A team that shares tips, mistakes and good examples gets better and safer, faster than people working alone.

Why learning together works

• People pick up practical tips from coworkers who do the same work.
• Mistakes become lessons instead of secrets.
• Less confident people get support without feeling judged.
• The team builds shared habits, such as always checking facts.

Simple ways to learn as a team

A worked example

Imagine the front office of a 25-person engineering firm. The office lead starts a "tip of the week" at the Monday meeting. Week one, a coordinator shows how she asks the approved tool to turn meeting notes into action items. Week two, a drafter shares how the tool once invented a code clause, and how he caught it. Within a month, most of the office is using the tool for routine drafting, and everyone knows the checking habit.

Supporting different comfort levels

Some people are excited about AI. Some are worried or sceptical. Both are reasonable. Good team habits include:

• Never shaming anyone for not knowing, or for being cautious.
• Respecting that some roles will use AI more than others.
• Keeping the focus on useful, safe tasks, not on showing off.
• Making sure frontline staff, not just office staff, are included.

Confidence comes from small, safe successes, repeated and shared.

You now know what AI is and is not, how chat tools work, why they can be wrong, how to use them well, and how to stay safe. The final step is turning that knowledge into habits. Small, steady practice beats a big burst of effort.

Your 30-day plan

The habits to keep

1. Use approved tools for work tasks.
2. Protect data: when in doubt, leave it out.
3. Prompt clearly: Task + Context + Format + Tone.
4. Check everything: read, verify, fix, own.
5. Pause on urgent requests and verify through a known channel.
6. Be fair and honest about how you use AI.
7. Ask and report early.

A worked example

Imagine Aisha, a shift supervisor at a food processing plant. In week one she finds the company's AI guideline and saves the IT contact. In week two she uses the approved tool to draft her shift handover notes each day. In week three she notices the tool sometimes rounds numbers, so she always checks counts against the log. In week four she shows two other supervisors her handover prompt. Thirty days in, she saves time every shift and trusts her own checking process.

Track your progress

• How many tasks did AI help with this month?
• How many mistakes did you catch?
• Did you share at least one tip?
• Do you feel more confident than when you started?

AI will keep changing. The habits in this course will stay useful: stay curious, stay careful and stay in charge. Thank you for taking the course.