Walkthrough · Responsible AI
5 Habits That Make You a Responsible AI User at Work
Build the daily reflexes that separate competent AI users from sloppy ones.
Most people reach for AI before they think, skip the spot check, and say nothing about how a piece of work was made. This lesson gives you five concrete habits that fix that pattern: think before you reach for the tool, verify every time, disclose, run a quick ethics check, and lead by example. Each habit takes seconds to apply and compounds into a professional reputation that AI alone cannot build.
Next step
What you will learn
- Apply a pre-task thinking routine before using any AI tool
- Spot-check AI outputs consistently before sharing them
- Decide when and how to disclose AI involvement in work
- Run a 30-second ethics check when AI touches people-related data
Lesson steps
Habit 1: Think Before You Reach for the Tool
AI is a means, not a goal. Know what you actually need before you open the tool.
The first habit is to think about the task before you reach for the tool. That sounds obvious, but most people open AI first and define the task second. That order is backwards.
Before you start, ask four questions: What does my manager actually need? What is the audience? What is the format? What is the constraint? Those four questions take 60 seconds and change everything about whether AI is the right move at all.
Once you know what you are actually trying to produce, you can make a real decision: reach for AI, or reach for something else. AI is a means, not a goal. The goal is the output your manager or your audience needs.
A journalist does not pick up the phone before knowing who they are calling or why. They draft the question first, then dial.
Classroom version: Before opening an AI tool to draft a summary for your director, write one sentence stating what the director needs to do or decide after reading it. That sentence is your real task. Now you know whether AI helps or not.
Try it: Before your next AI task, write answers to all four questions on a sticky note or in a blank doc. Then decide: does AI actually help here?
Think first, tool second. AI is a means, not a goal.
Habit 2: Verify Every Time
AI is fast and confident, but that does not mean it is always right. Spot check every number and claim.
The second habit is to verify, and the key phrase is every time. AI is fast and confident. That combination is dangerous because confidence signals accuracy even when it does not deliver it. An output that reads like a polished report can still contain a wrong number, a fabricated citation, or a mischaracterized claim.
The specific standard here is this: the five seconds it takes to spot check a number or a claim is the difference between a competent professional and a sloppy one. That is not a metaphor. Five seconds. Check the number against the source. Read the claim against the context. If you cannot verify it in the time it takes to read it twice, that is a signal to dig further before you share it.
Verification is not a sign that you distrust AI. It is a sign that you take ownership of the work that leaves your name. AI produces the draft; you are responsible for the final output.
An accountant does not file a return because the software calculated it. They review the figures against the source documents before signing.
Classroom version: You use AI to generate market size figures for a slide deck. Before the deck goes to leadership, open the original report or data source and confirm at least the three most prominent numbers. This takes under two minutes and is the line between competent and sloppy.
Try it: On your next AI-assisted document, highlight every number and named claim. Verify each one before you send it. Time yourself. It will be under five minutes total.
Verify every time. Five seconds of checking is the line between a competent professional and a sloppy one.
Habit 3: Disclose
Disclose AI involvement not because it is always required, but because it builds trust and the alternative is much worse.
The third habit is to disclose when you have used AI in your work. The reason is not primarily compliance. The speaker is clear about this: disclose not because it is always required, but because it builds trust.
There is also a practical risk calculation built into this habit. The alternative to disclosing is being caught not disclosing. And being caught not disclosing is much worse than the disclosure itself. In almost every professional context, the damage from concealment is larger than the awkwardness of transparency.
Disclosure does not have to be formal or lengthy. It can be a single sentence in an email, a note in a document footer, or a verbal comment in a meeting. The form matters less than the consistency. Make it a default, not a decision you re-litigate every time.
A ghostwriter who is honest with a client about their process keeps the client's trust. The one who hides the process risks losing the relationship the moment it surfaces.
Classroom version: You draft a client-facing proposal using AI to structure the sections. At the bottom of the document, add one line: 'Initial draft structure generated with AI assistance, reviewed and edited by [your name].' That sentence costs you nothing and protects your credibility.
Try it: Draft a one-sentence disclosure statement you can drop into any document or email where AI contributed. Keep it in a note you can copy from quickly.
Disclose by default. Being caught not disclosing is much worse than the disclosure itself.
Habit 4: Think About Ethics Every Time
Every time AI touches people-related data, run a quick 30-second ethics check. No exceptions.
The fourth habit is to think about ethics every time you use AI for something with people in it. The phrase every time appears deliberately. This is not a one-time policy review or an annual training module. It is a reflex you build into each individual task.
The specific categories the speaker names are: customer data, employee data, hiring, and evaluation. These are the areas where AI can produce or reinforce bias, expose private information, or make consequential decisions that affect real people without adequate human judgment in the loop.
The practical tool is a quick ethics check that takes 30 seconds. Before you run that AI process or share that AI-generated output, pause and ask: Is there personal data involved? Could this disadvantage a group? Am I the right person to make this call, or does it need a second set of eyes? Thirty seconds is not a burden. It is a minimum.
A pharmacist checks for drug interactions before dispensing every prescription, not just the complicated ones. The habit applies universally because the stakes apply universally.
Classroom version: You are using AI to help screen job applications. Before running the tool, spend 30 seconds asking: What data is going in? What criteria is the model using? Who reviews the output before any decision is made? If you cannot answer all three, that is your signal to slow down.
Try it: Write three ethics-check questions on a card and keep it visible at your desk. Ask all three before any AI task that involves data about people.
Run the ethics check every time. Thirty seconds is all it takes when AI touches customer data, employee data, hiring, or evaluation.
Habit 5: Lead
Even as a junior team member, the way you handle AI sets norms for the people around you.
The fifth habit is to lead. This one surprises people because leadership feels like a title, not a behavior. The speaker is direct: even if you are not in a leadership role today, even if you are a junior on a team, the way you handle AI sets norms for the people around you.
Norms are not set by policy documents. They are set by what people see their colleagues doing. When you consistently think before reaching for the tool, verify outputs, disclose AI involvement, and run ethics checks, you make those behaviors visible. Other people notice. Some of them copy you. That is how a team culture actually forms around responsible AI use.
You do not need permission to lead on this. You just need to do it consistently and visibly enough that it registers. The five habits in this lesson are not just personal practices. They are the behaviors that signal to everyone around you what the standard looks like.
A new employee who arrives early and prepares thoroughly sets a quiet standard that teammates feel even before anyone names it. They never announced a policy. They just showed what the bar looks like.
Classroom version: In your next team meeting where AI output is presented, be the person who asks 'How did we verify this?' That one question, asked consistently, becomes the norm. You do not need a title to ask it.
Try it: Choose one of the five habits to practice visibly in a team setting this week. Name it out loud once when you use it, so colleagues see the behavior, not just the output.
Lead through behavior. The way you handle AI sets norms for the people around you, no title required.
Transcript
- 0:00 Let's start with the first habit.
- 0:03 Think about the task before you reach for the tool.
- 0:07 AI is a means, not a goal.
- 0:09 What does my manager actually need?
- 0:12 What is the audience?
- 0:15 What is the format?
- 0:16 What is the constraint?
- 0:17 Once you know that, you reach for AI
- 0:20 or reach for something else.
- 0:23 The second habit is verify.
- 0:26 Every time.
- 0:28 AI is fast and confident,
- 0:31 but that does not mean it's always right.
- 0:33 The five seconds it takes for you to spot check a number,
- 0:37 a claim, that is the difference
- 0:40 between a competent professional and a sloppy one.
- 0:44 The third habit is disclose.
- 0:47 Not because it's always required,
- 0:49 but because it builds trust.
- 0:52 And because the alternative, being caught not disclosing,
- 0:57 is much worse.
- 0:59 The fourth habit is think about ethics every time.
- 1:05 Every time you use AI for something with people in it,
- 1:08 customer data, employee data, hiring, evaluation,
- 1:13 run a quick ethics check.
- 1:15 It takes 30 seconds.
- 1:17 The fifth habit is lead.
- 1:20 Even if you are not in a leadership role today,
- 1:24 even if you are a junior on a team,
- 1:26 the way you handle AI sets norms for the people around you.
Questions
Do I need to disclose every time I use AI, even for small tasks?
The habit is to default to disclosure, not to disclose only when required. The speaker frames it this way: disclose not because it is always required, but because it builds trust. And because being caught not disclosing is much worse than the disclosure itself. A short, consistent disclosure costs little and protects a lot.
What counts as an ethics check? Is there a specific process?
The speaker describes it as a quick check that takes 30 seconds. The specific trigger is any AI use that involves people: customer data, employee data, hiring, or evaluation. The check does not require a formal tool. The goal is to pause and ask whether the use is appropriate before proceeding, not to complete a form.
What if my team does not follow these habits? Can I really lead without a title?
Yes. The speaker is explicit: even if you are a junior on a team, the way you handle AI sets norms for the people around you. Norms form through visible behavior, not through policy. Doing these habits consistently and in front of colleagues is itself a form of leadership.
How do I know when to use AI versus something else?
Answer the four pre-task questions first: What does my manager actually need? What is the audience? What is the format? What is the constraint? Once you know that, you reach for AI or reach for something else. AI is a means, not a goal. The decision follows from the task definition, not the other way around.
Glossary
- Spot check
- A quick, targeted verification of a specific number or claim in an AI output, performed before sharing the work. The speaker sets the standard at five seconds per item.
- Ethics check
- A 30-second pause before any AI task involving people-related data (customer data, employee data, hiring, evaluation) to assess whether the use is appropriate and fair.
- Disclosure
- Informing a manager, client, or audience that AI was involved in producing a piece of work. Done by default to build trust, not only when formally required.
- Norm-setting
- The process by which visible, consistent behavior by one team member becomes the implicit standard others follow, regardless of titles or formal policy.
Resources
- AI Fundamentals for Teams Foundational course that supports all five habits with deeper context on how AI tools work and where they fail
- Responsible AI Checklist A printable pre-task checklist combining the four questions from Habit 1 and the ethics check from Habit 4