Lesson 3: AI and Your Job - What to Tell Your Boss
On May 2, 2023, Chegg's stock dropped 48% in a single day. The education technology company told investors that ChatGPT was impacting their business. Students were using free AI instead of paying for Chegg's tutoring services.
The CEO admitted they had "seen a significant spike in student interest in ChatGPT" and it was affecting new customer growth.
Within hours, nearly $15 billion in market value disappeared.
Here's the uncomfortable truth: AI is changing jobs, and how you navigate that change will affect your career. Use it wrong and you could get fired. Use it right but talk about it wrong, same result. Don't use it at all? You might get outpaced by someone who does.
You need a strategy for AI at work that protects your job while helping you do it better.
The Transparency Spectrum
There's no one-size-fits-all answer for whether to tell your boss you're using AI. It depends on your company, your role, and what you're using it for.
Always Disclose
Client-facing work: If you're a lawyer, consultant, accountant, or anyone who bills hours or provides professional services, you probably need to disclose AI use. Some clients explicitly ban it in contracts. Others require disclosure for ethical reasons.
Regulated professions: If you're in healthcare, finance, legal, or another regulated field, AI use might require disclosure or approval. Check your professional guidelines.
Content you're claiming as original: If you're submitting work for a grade, publication, or attribution, and you used AI substantially, you likely need to disclose it. Academic institutions are creating policies around this. Err on the side of transparency.
When your company explicitly asks: Some companies are surveying employees about AI use or updating policies. If they ask, tell the truth.
Consider Disclosing
Team projects: If using AI gives you a significant speed advantage and others might wonder how you're working so fast, consider sharing your process. This can position you as an innovator rather than someone who looks suspiciously efficient.
Process improvements: If you figured out an AI workflow that could help your whole team, sharing it can make you look like a leader. Frame it as "I found a way to cut our report writing time in half."
When you want official approval: If you're planning to use AI extensively for a major project, getting explicit buy-in from your manager protects you if questions come up later.
Keep Quiet (For Now)
Personal productivity: If you're using AI to write better emails, organize your notes, or understand complex topics faster, this is like using a spell-checker or search engine. It's a tool for your own efficiency.
Drafting and brainstorming: If you use AI to create a first draft that you then heavily edit, or to brainstorm approaches you then research yourself, you're still doing the thinking. The AI is just speeding up the "getting started" phase.
In companies that haven't set a policy: If your company hasn't said anything about AI one way or another, using it for general productivity is probably fine. But don't assume silence means approval for high-stakes use.
When your industry is in flux: Some industries are still figuring out what AI means for them. If you're in one of those, keep your head down until clearer guidance emerges.
Companies That Banned AI (Then Changed Their Minds)
JP Morgan Chase banned employees from using ChatGPT in early 2023, citing security and compliance concerns. By mid-2024, they had developed their own internal AI tools and were actively encouraging employees to use them for productivity.
The lesson: companies often ban first and figure out controlled use later.
Apple initially banned ChatGPT over fears that employees might leak proprietary information. Then they announced they were integrating AI features across their product line.
Amazon restricted ChatGPT use, then launched their own AI assistant (Q) for employees.
The pattern is clear: large companies are moving from "ban everything" to "control access." If your company currently bans AI, that might not be permanent. But respect the current policy or you could get fired for violating it.
How to Pitch AI Use to Your Boss
If you want official approval to use AI for work, here's how to frame the conversation:
Bad approach: "I've been using ChatGPT for months and it's amazing! You should let everyone use it."
This makes your boss worry about what you've already exposed and puts them on the defensive.
Better approach: "I've noticed AI tools are becoming popular in [industry]. I'd like to explore whether they could help our team with [specific task] while staying within our security policies. Can we discuss what would be appropriate?"
This positions you as thoughtful and aware of risks, not reckless.
The Script
"I've been reading about teams in [industry] using AI to speed up [specific task that takes your team significant time]. I think there might be an opportunity for us to do similar work faster while maintaining quality. Would you be open to me researching secure options and proposing a small pilot project? I want to make sure we do this in a way that fits our company's policies and protects sensitive information."
What Your Boss Is Worried About
When you bring up AI, here's what's running through your manager's mind:
- Will this expose confidential data? (Answer: explain how you'll avoid that)
- Will this make jobs redundant? (Answer: frame it as "doing more with the same team")
- Will this create liability if the AI screws up? (Answer: emphasize human review of all output)
- Will I get in trouble with my boss for allowing this? (Answer: suggest starting with a small, controlled test)
Address these fears proactively and you'll have a much better chance of getting approval.
The "Invisible Efficiency" Trap
Here's a counterintuitive problem: if AI makes you too efficient too fast, it can actually hurt your career.
Let's say you normally take a week to write a report. With AI, you can do it in a day. If you turn it in a day later, your boss might think:
- The work must not be thorough
- You were slacking off before
- The job doesn't need a full-time person
The solution isn't to stop using AI. It's to use the extra time strategically.
Deliver the report in 4 days instead of 1. Use the extra three days to:
- Make the report more thorough
- Take on an additional project
- Invest in learning something new
- Help a colleague with their work
Your output quality and quantity should increase gradually, not suddenly. This looks like professional growth, not like you were sandbagging before.
How to Document Your AI-Assisted Work
If you're disclosing AI use (or think you might need to later), document what you did:
Template:
Project: [name]
AI Tool Used: [ChatGPT/Claude/etc., including version if known]
Date: [when you used it]
Purpose: [what you used it for]
Prompts: [general description, or save actual prompts]
Output: [what the AI produced]
Human Review: [what you checked, changed, or verified]
Final Deliverable: [what you actually submitted]
You don't need this for every email you write with AI help. But for substantial projects, having this record protects you if questions come up later.
Save this documentation somewhere secure. If there's ever a dispute about your work, this shows you were thoughtful and transparent about your process.
When AI Use Must Be Disclosed
Some situations have clear rules:
Academic work: Most universities now have AI policies. Some allow it with disclosure, some ban it entirely, some treat it like any other reference tool. Check your institution's policy.
Licensed professions: The American Bar Association and other professional organizations are developing guidelines. If you're licensed (lawyer, CPA, doctor, etc.), check your professional body's latest guidance.
Published work: If you're submitting articles, books, or research for publication, check the publication's AI policy. Many now require disclosure if AI was used substantially in creating the work.
Legal proceedings: If you're involved in litigation and used AI to prepare filings or evidence, you likely need to disclose that. Courts are still developing standards here, but transparency is the safer path.
Career Protection Checklist
Before using AI for any work task:
- [ ] Have I checked if my company has an AI use policy?
- [ ] Does my contract or client agreement restrict AI use?
- [ ] Am I in a regulated profession with specific AI guidelines?
- [ ] Will I need to defend this work's originality or accuracy later?
- [ ] Could this expose confidential or proprietary information?
- [ ] Would I feel comfortable explaining this use to my boss?
- [ ] Am I prepared to verify any facts or claims the AI generates?
- [ ] Have I saved documentation of how I used AI (if needed)?
If you answered "no" or "I'm not sure" to any of these, slow down and get clarity before proceeding.
The Bottom Line
AI isn't going to take your job. But someone using AI might.
The goal is to use these tools to increase your value without increasing your risk. That means being smart about what you use AI for, how you talk about it, and when you keep quiet.
Transparency isn't about confession. It's about professionalism. Use AI in ways you're comfortable defending, document what you did, and understand your company's or industry's expectations.
The worst position you can be in is having to explain AI use you didn't think would come to light. The best position is having used AI strategically, documented it appropriately, and having work results that speak for themselves.
Your job is safe if you're delivering value. AI should be your tool for delivering more of it, not your secret shortcut that could blow up later.