What Not to Put in ChatGPT When Writing Your Resume

AI Resume Writing is Dangerous

I’m a professional resume writer. I’m not anti-AI. In fact, I honestly think tools like Gemini or ChatGPT can be extremely valuable when enhancing your career materials.

But the way most job seekers use them is dangerous.

I see people pasting their entire lives—unfiltered—into public chatbots. They hit “Enter,” and suddenly their home address, their manager’s private email, and their company’s internal revenue data become potential training data for a global model.

If you are wondering what not to put in ChatGPT, the short answer is: anything that identifies you, your boss, or your company’s trade secrets.

Be Careful What You Tell AI

As a human writer, I am bound by privacy ethics and confidentiality agreements. Public AI models are not.

A recent study led by Dr. Jennifer King at the Stanford Institute for Human-Centered AI analyzed the privacy policies of major AI tools including ChatGPT, Gemini, Claude, Nova, MetaAI, and Copilot. The research found that ALL of them appear to use customer chats to “train and improve their models by default” and that “some retain this data indefinitely.”

This means if you paste your raw resume into a free chatbot, you are potentially exposing sensitive information to the public domain.

What Not to Put in ChatGPT

To protect your privacy, here are the three things you should never enter into a public AI chatbot:

  1. Personal Details: Your name, home address, personal phone number, and personal email do not belong in a chat window. This is cybersecurity 101.
  2. Reference Contact Info: I see this constantly. Candidates upload their reference sheet to “fix the formatting.” You are essentially doxxing your former employers by uploading their direct phone numbers and emails to a public database without their consent.
  3. Internal Data (that isn’t public): Redact specific revenue details, business reports, client data, source code, or anything protected by a non-disclosure agreement. “Managed a $5M budget” is fine. “Generated $5M in revenue for [Client Name] via [Secret Internal Strategy]” is a breach of your NDA. If it wasn’t in a press release, DON’T feed it to AI.

The “Safe” Way to Prompt (The Bracket Method)

You don’t have to stop using AI to optimize your resume. You just have to be smart about HOW you use it. Before you prompt your AI assistant of choice, take 2 minutes to “sanitize” your resume using brackets:

Instead of “Jane Doe, 123 Maple St.” write: “[Name], [Address]”

Instead of “Project for Coca-Cola” write: “Project for [Major Beverage Client]”

The AI doesn’t need to know you worked at Google to fix your grammar; it just needs to know you worked at a Large Tech Company.

Human vs. AI Resume Writing

AI is fantastic for speed, grammar, and brainstorming. However, it often lacks nuance, discretion, and the ability to tell a cohesive career story.

If you want a human eye on your resume, Red Pen Wench can help. We offer 30-minute and 1-hour job search consultations. In a 1:1 video call, we can discuss your career goals, revise your career materials, and sharpen your narrative in a safe, confidential environment. Book a Consultation Here.

–Jillian Kinsey, Owner & Founder of Red Pen Wench