You've been staring at the same email for five minutes. A client is unhappy. You know what you want to say. But every version sounds wrong — too defensive, too apologetic, too something.
So you open ChatGPT, paste in the client's message, type "write a professional response to this complaint," wait for the output — and immediately delete it. Because it reads like it was written by a corporate chatbot with no personality.
The problem isn't AI. The problem is how most people use AI for email.
When AI actually helps — and when it doesn't
AI is not a ghostwriter. If you treat it like one, you get bland, generic text that could've been written by anyone. And people notice. We've all developed a gut feeling for emails that are "too polished" — the ones that feel like they were generated rather than written.
Here's where AI genuinely earns its keep:
Turning rough notes into real emails. You jot down three bullet points of what you want to say. AI turns it into a coherent message. Your thoughts, better wording.
Adjusting tone. You wrote something, but it's too casual for a client. Or too stiff for a colleague. AI can shift the tone without changing what you're actually saying.
Writing in a second language. You're drafting in English but it's not your first language. AI smooths out the rough edges without changing your meaning.
Cutting the fluff. You wrote three paragraphs where one would do. AI helps you get to the point.
Where AI does not help: when you haven't figured out what you actually want to say. No tool can do your thinking for you. You need to sort out the substance first.
The mistakes everyone makes
Mistake 1: Outsourcing the whole email. You type a vague prompt — "Write an email to my boss about next week" — and hope for magic. What you get is interchangeable filler that sounds like it came from a template.
Mistake 2: Not editing the output. AI gives you a draft. A draft. Not the finished email. If you hit send without reading it carefully, you just sent something that doesn't sound like you. People can tell.
Mistake 3: Giving bad instructions. "Make it more professional" is not a useful prompt. "Make it friendly but firm — I need to hold them to a deadline" is.
Mistake 4: Using AI for the wrong emails. A condolence message? A genuine apology? A sensitive conflict? Those emails need to come from you. Actually from you.
How to keep your own voice
The most important rule: always write first yourself. Even if it's just bullet points. Even if the draft is rough. When you start with your own words, your voice stays in the text — even after AI has polished it.
An example. You need to turn down a project because you're at capacity. Your first draft:
Hey Sarah, thanks for reaching out. This sounds like a great project but I'm totally swamped right now and wouldn't be able to give it the attention it deserves. Could we reconnect in April?
Not bad. But you're not sure if the tone is right for this particular client. So you ask AI to refine it — "more professional but still warm":
Hi Sarah, thank you for thinking of me — the project sounds really interesting. Unfortunately, I'm fully booked at the moment and wouldn't want to take it on without being able to give it proper attention. Would it work to revisit this in early April?
Same content. Your tone. A bit more polished. That's the sweet spot.
The real problem: context switching
Here's the thing nobody talks about. How many people actually copy their emails into ChatGPT, get a rewrite, and paste it back? In practice, almost nobody does this consistently. Not because it doesn't help — but because it's too much friction.
You're writing an email in Outlook. To use AI, you have to:
- Select and copy the text
- Switch to your browser
- Open ChatGPT
- Write a prompt and paste your text
- Wait for the response
- Read the result, copy it
- Switch back to your email client
- Paste it in, delete the old text
Eight steps. For a writing assist.
No wonder most people just send the email as-is. Or spend five minutes rewording something themselves that AI could've handled in three seconds.
This is why we built Schriftly. It's a desktop app for Windows and macOS that works inside any application — Gmail in your browser, Outlook, Apple Mail, Slack, wherever you write.
You select your text, press Option+P (Mac) or Ctrl+Shift+P (Windows), and get a polished version instantly. No tab switching. No copy-pasting. No waiting for a chat response.
It sounds like a small difference. But that small difference is what determines whether you actually use AI day-to-day — or just think it's a nice idea in theory.
Practical tips for better AI-assisted emails
A few things that actually work in practice:
Start with bullet points. If you're stuck, write three quick bullets of what you need to say. Let AI shape them into an email.
Always read the output. Take ten seconds to skim the result. Would you say it that way? If not, tweak it.
Use AI for the boring emails. The meeting confirmation, the out-of-office reply, the quick "thanks for the update" message — AI is perfect for these. Save your energy for the emails that matter.
Be specific with your instructions. Instead of "make it better," try "cut this in half and make the tone more direct." The more precise you are, the better the result.
Don't forget the subject line. A good subject line decides whether your email gets opened. AI can help here too — especially when you're drawing a blank.
The short version
AI won't make you a better writer. But it will make writing faster and smoother — if you use it right.
Write first, let AI refine. Not the other way around.
And make sure AI is where you need it: right inside the app you're already writing in. Not three tab switches away.