Is it cheating to use AI? Honest answers for school, work, and personal life

If you've ever felt a flicker of guilt asking ChatGPT to write an email or help with homework, you're not alone. AI has scrambled the rules around what counts as your own work. Here's the honest take, broken down by situation.

At school: where the line is

What is usually fine

• Using AI to understand a concept that confused you in class
• Brainstorming essay topics
• Asking AI to suggest where your draft is weak (then you fix it)
• Using AI as a tutor: "Quiz me on the French Revolution."
• Spell-check, grammar check, basic editing of your own writing
• Asking AI to summarize a textbook chapter so you can study faster

What is usually cheating

• Pasting the assignment prompt into AI and submitting the response as your work
• Having AI write your essay, then making cosmetic edits
• Using AI on a test or quiz that prohibits it
• Having AI solve math problems and copying answers without learning the method
• Having AI complete code assignments for a programming class

The gray area

• Letting AI suggest sentence rewrites, then accepting them: depends on assignment
• Using AI to translate your draft to better English: depends on the class
• Asking AI to "explain this answer to me" for homework you got wrong: usually fine

What to actually do

1. Find your school's AI policy. Many have one now.
2. If your teacher hasn't said anything, ask. "Are we allowed to use ChatGPT to help brainstorm for this essay?"
3. Default to citation if you're unsure. "I used ChatGPT to suggest the structure of my argument."
4. Whatever you submit, make sure you could defend it in conversation. If your teacher asked "tell me what you wrote and why," could you?

At work: where the line is

What is usually fine

• Drafting emails (the most common AI use at work)
• Summarizing long documents or meeting notes
• Brainstorming ideas for a project
• Cleaning up a slide deck or report
• Generating images for internal presentations
• Coding assistance for technical roles

What is usually a problem

• Pasting confidential data into public ChatGPT. Customer lists, financial data, legal documents, employee info. Companies have lost data this way. Many companies now block public ChatGPT entirely.
• Submitting AI-written work as expert analysis. If your job is to analyze, you need to actually analyze. AI is your assistant, not your replacement.
• Using AI for things requiring real human judgment. Hiring decisions, performance reviews, legal opinions.
• Pretending you did work you didn't do. If your boss thinks you wrote a 30-page report and AI wrote 28 of them, that's a problem.

What to actually do

1. Check if your company has an AI policy. Most do now.
2. If your company has approved AI tools (Copilot for Microsoft 365, ChatGPT Enterprise), use those for work. They have privacy and security in place.
3. Never paste customer data, financial data, or anything confidential into a personal AI account.
4. Treat AI like a smart intern. They draft; you review and own the final.
5. If you're uncertain about a specific use, ask your manager.

In personal life: there are no rules

For your own emails, your own planning, your own brainstorming, your own creative projects, AI is just a tool. Use it freely:

• Drafting birthday card messages
• Writing a complaint letter to your landlord
• Planning a trip
• Coming up with names for a pet
• Helping you understand a confusing letter
• Practicing a difficult conversation
• Generating art for fun

The only personal-life rules:

• Don't pretend AI work is human work in ways that could mislead (a "personal" letter to a loved one is best when it's actually from you).
• Don't paste sensitive personal info (SSN, full account numbers) into public AI.

The "could you defend it?" test

One useful test across all situations: would you be comfortable explaining exactly how you made the thing to whoever is reading it?

• For a school essay: "Could I defend every claim in conversation?" If yes, you're fine. If no, you went too far.
• For a work email: "Could I have written this myself given enough time?" If yes, AI just made it faster. If no, you may be in over your head.
• For a personal letter: "Does this still sound like me?" If yes, you're fine. If no, edit until it does.

The "told them, told them again" rule

If you're using AI in a context where the recipient might care, disclose. Examples:

• Submitting school work: follow your school's disclosure rules.
• Publishing a blog or article: increasingly expected to disclose AI assistance.
• Sending a "personal" letter: if AI wrote most of it, it's not really a personal letter.

If they don't care or won't notice (a thank-you email, a quick reply to a customer service question), no disclosure needed.

"But isn't using a calculator different from using AI?"

Same kind of objection that came up with calculators in math class, spell check, Google for research, GPS for navigation. The pattern with each new tool:

1. Initially banned or considered cheating.
2. Mixed acceptance: some teachers/jobs ban it, others embrace it.
3. Universal acceptance with rules for when it's appropriate.

AI is somewhere between step 1 and step 2 right now. The norms are settling. Don't assume yesterday's rules apply, and don't assume next year's rules apply either.

What is changing

• Many universities now allow AI for brainstorming and editing but ban it for first drafts.
• Some employers list AI proficiency as a job requirement.
• Other employers run AI-detection software on submitted work.
• Many publishers require AI disclosure on submissions.
• Most workplaces now have explicit AI policies.

Three questions to ask yourself

1. What is the purpose of the assignment? If it's to demonstrate your skill, AI should support that. If it's to produce a useful output, AI is fine.
2. Could I do this without AI given enough time? If yes, AI is making you faster. If no, you may be misrepresenting yourself.
3. Would I be embarrassed if the recipient knew exactly how I made this? If yes, that's your conscience speaking. Adjust.

The bottom line

AI is a tool. Using a tool isn't cheating. Hiding the use of a tool to mislead someone about your work is the problem. For most uses, AI is no different from spell check or Google. For school and high-stakes work, follow the rules and use your judgment. For personal life, use it freely and enjoy it.