"I used ChatGPT": why that line hurts and what to say instead

Some freelancers think honesty means opening with:

I used ChatGPT to write this proposal.

It feels transparent. It often backfires. The client did not hire a language model. They hired you. That line trains them to doubt every sentence that follows, including the parts that are actually strong.

This article explains why the confession fails, what buyers assume when they read it, and what to say instead if you use AI in your workflow. For a full process section on tools, see how to mention AI tools in a proposal. For better first lines overall, see freelance proposal opening lines.

What the client hears when you lead with ChatGPT

They rarely think, “Great, this person is efficient.” More often they think:

• This proposal might be generic (same template as fifty others)
• They did not think hard about my job (the model did)
• Quality might be sloppy (no pride in craft)
• They might paste my secrets into a tool (data worry)
• They are inexperienced (oversharing instead of selling competence)

You might intend humility. They read risk.

Risk is the enemy on marketplaces and in inboxes. Why clients ignore proposals applies here: you added a red flag in line one.

Why transparency in the wrong place feels like low effort

Transparency works when it answers a client question:

• How do you handle confidentiality?
• How do you ensure accuracy?
• What is your review process?

Transparency fails when it sounds like a disclaimer for work you have not done yet:

I used AI, so if something is wrong, that is why.

That shifts responsibility away from you before you earned trust.

Professional freelancers can use assistants and still own the deliverable. Ownership is what you should signal in line one, not which button you clicked.

The difference between tool mention and tool apology

Apology (weak):

I used ChatGPT, hope that is okay.

Process (stronger):

I draft faster with modern tools, then edit everything against your brief before you see it.

Unnecessary in opening (usually):

This message was AI-generated.

If the whole proposal reads like a template, admitting AI does not fix it. Specificity fixes it: their goal, their constraint, your plan.

What to write instead: openings that work

Replace the ChatGPT line with an outcome-first opening. Pick the pattern that fits the job.

Pattern 1: Mirror their outcome

You need [specific outcome] by [constraint]. I would start by [sensible first step] so you can review direction before we commit to the full scope.

No tools named. Pure competence.

Pattern 2: Show you read the post

Your note about [detail from post] matters because [short reason]. Here is how I would approach it without delaying [deadline/milestone].

This is the fastest way to prove you are not a blind paste.

Pattern 3: Risk-aware professional

The main risk in projects like this is [friction]. I would address it by [concrete action] in milestone 1.

Works for technical, CRO, and operations gigs.

Pattern 4: Small safe first step

I suggest starting with [paid milestone] so you can approve [artifact] before the larger build.

Especially strong for beginners; see beginner freelancer proposals without big case studies.

Pattern 5: When they explicitly asked about AI

Put the answer in the second paragraph, not the first:

You asked whether I use AI. I use assisted drafting for speed, with human editing and fact-checking before delivery. For your project, that means [benefit to them: faster variants, quicker research summaries, etc.].

Lead with their question’s business impact, not your tool stack.

Before and after examples

Web development post

Before:

I used ChatGPT to help write this. I can build your site.

After:

You need a marketing site migrated to [stack] with forms connected to [CRM]. I would begin with a short audit of your current pages and plugin list, then propose a fixed milestone for the homepage template before rolling out the rest.

Copywriting post

Before:

Full disclosure: ChatGPT helped with this cover letter.

After:

You need landing copy that explains [offer] to [audience] without sounding corporate. I would draft a section map for your approval, then write hero and proof sections first so you can check tone early.

VA or ops post

Before:

AI wrote this but I can do the tasks.

After:

You need inbox triage and weekly reporting by Friday. I would set up labels and a one-page SOP in week one so you can see the system before we expand tasks.

Notice the after versions never hide tools forever. They just do not lead with a warning label.

Where AI belongs if you mention it at all

Put tool talk under Process, Quality, or Tools near the middle, after:

1. You restated the job
2. You outlined approach or milestones
3. You showed one proof point

Example mid-proposal line:

I use assisted research to gather competitor angles quickly; final strategy and copy are edited manually to match your voice.

One sentence. Move on.

If you already sent the bad line

Do not panic. Your follow-up can recover partially:

Quick clarification on my proposal: the important part is the plan in paragraph two. I use drafting tools for speed, but I personally review every deliverable. Happy to start with a small milestone on [specific artifact] so you can judge the work directly.

Short. No drama. Offer a tangible next step.

For more follow-up patterns, use first reply after client messages you if they responded with skepticism.

How to keep AI assistance without sounding AI-written

Read the proposal aloud. If you stumble, a client will too.

Add one post-specific sentence they know you could not guess without reading.

Cut filler: “delve,” “robust,” “leverage,” “in today’s digital landscape.”

Name a tradeoff: what you will not do in v1.

End with one clear action.

Run the proposal checklist before resending anything.

Myths freelancers tell themselves

“Clients want full honesty about AI upfront.” They want honesty about quality and data. Tool brand in line one is rarely the right hook.

“Everyone uses ChatGPT anyway, so it is fine to say it.” Everyone eats lunch too; you do not open a proposal with that. Compete on clarity and scope.

“If I do not admit AI, I am lying.” Using assistance is not lying. Delivering unedited garbage while claiming expert craft is lying. Focus messaging on review and ownership.

“Admitting AI explains why my English is perfect.” Non-native speakers sometimes fear they sound “too polished.” You do not need to confess AI for that. See non-native English proposal mistakes for natural tone without self-sabotage.

When you should say you do not use AI

If the post says no AI, academic integrity rules apply, or the client fears plagiarism, state clearly:

I do not use generative AI on your deliverables for this project.

That is a selling point, not a confession.

Checklist: remove the hurt line, add trust

• Delete “I used ChatGPT” from the opening
• First sentence is about their outcome or constraint
• At least one detail from their post appears in paragraph one or two
• One proof point or relevant experience, honestly labeled
• Optional single process line on tools later, with review and data habits
• Clear next step (milestone, questions, or call window)
• Read aloud; remove generic AI filler phrases

What this is not

This is not advice to deceive clients about automation. It is advice to stop leading with the wrong signal. You are not selling ChatGPT. You are selling a reliable human who uses modern workflows and still signs their work.

Bottom line

“I used ChatGPT” in line one tells the client to doubt you before they meet you. Lead with understanding, plan, and low-risk next steps. Mention tools only where they support trust: process, confidentiality, and quality control.

Your proposal should sound like someone who already started thinking about their job, not someone asking forgiveness for how they typed.

Fix the opening, keep the honesty where it helps, and let the work you describe do the selling.