How can marketing teams use ChatGPT, Copilot and Claude without sounding like everyone else?

Brief the tool on your brand before you ask it for a single word, use it as much for the work that surrounds the writing as for the writing itself, and have a person edit and fact-check everything before it goes out. AI copy sounds like everyone else's when it gets the same brief as everyone else's. Give ChatGPT, Copilot or Claude your voice, your audience and your proof, and the first draft starts from your brand instead of from nothing.

Here's how a marketing team puts that into practice, from the voice brief to the sign-off.

Why does AI marketing copy all sound the same?

Because the prompt is the same. Ask for "a LinkedIn post about our new service" and the tool has nothing to go on but the words in that sentence. It doesn't know who your customers are, how you talk to them, which phrases you'd never use or what you can prove that a competitor can't. So it fills the gaps with safe, familiar wording that would fit any business in your sector.

The fix isn't a cleverer tool. It's a better brief. The difference between a generic draft and one that sounds like you is the context you hand over at the start.

How do you brief AI with your brand voice?

Write a voice brief once and use it every time. Keep it short enough that people will actually maintain it. It needs:

• Who you're writing for, and what they care about before they buy.
• How you sound, in a few plain sentences, with words you use and words you avoid.
• Three or four examples of your best real work: an email that performed, a post that got the right replies, a page you're proud of.
• What makes you different, with the proof behind it.

Then store it where the tool can find it, so nobody pastes it in from memory. OpenAI's help article on projects in ChatGPT, checked in October 2026, says "Projects keep related chats, files, and instructions together", and that project instructions override your global custom instructions inside that project. Claude has the same idea. Anthropic's help centre in 2026 describes projects as "self-contained workspaces with their own chat histories and knowledge bases", with project instructions to shape Claude's responses. Set up one project per brand or client, load the voice brief and your examples, and every chat in it starts from your brand.

If your team works in Microsoft 365 with Copilot, keep the voice brief as a shared document and make pasting it in the first step of any writing task. The habit matters more than the tool.

Where does AI save a marketing team the most time?

Start with research, because it's where the generic problem gets solved. Paste in customer reviews, survey comments or sales call notes with personal details removed, and ask the tool to group them into themes and pull out the exact phrases customers use. Those phrases are worth more than any adjective a tool can suggest, because they're how your buyers describe their own problem. Competitor research works the same way: give it the pages you've collected and ask what each one promises and what nobody is saying.

First drafts come next. A good draft prompt names the angle, the one proof point, the reader and the action you want them to take, then points to the voice brief. Ask for three openings rather than one finished piece, pick the strongest and build from there. Your writer stays the writer. The tool clears the blank page.

Repurposing is where the hours come back. One article, webinar or case study becomes an email, a set of social posts and several ad variations to test. Brief each channel separately, because a LinkedIn post and a short ad headline are different jobs, and keep every version tied to the same facts so nothing drifts in the retelling.

Reporting closes the loop. Export the month's figures, paste them in and ask for a plain summary of what changed, what didn't and which questions the numbers raise. Every figure in the final report gets checked against the platform it came from, because a summary that misreads one column turns into a decision based on the wrong number.

What must a person do before anything goes out?

Edit for voice, then check every fact. Editing means cutting the stock phrases, adding the specific detail only your team knows and reading it aloud. If it could have been written by any business in your sector, it isn't finished.

Fact-checking matters even more, because a wrong answer from an AI tool reads exactly like a right one. A made-up statistic, a feature you don't offer or last month's price sits in the draft looking as solid as the truth. The advertising rules don't bend for that. The Advertising Standards Authority said in 2025 that "the same rules apply in the same way, regardless of whether AI is used or not". The same article confirms there's no blanket legal requirement in the UK to disclose that AI was used in an ad, so the question isn't whether a tool helped. It's whether every claim in the ad is true and can be backed up. A claim the AI invented is still your claim.

Give every piece a named person who signs it off. That's the person who checked the facts, the prices and the promises, and who stands behind them.

What data should never go into an AI tool?

Your data protection lead has the final word on this, and what follows is general information rather than legal advice.

Customer lists, contact details and anything else that identifies a person stay out of any tool your organisation hasn't approved for that data. So do unreleased campaigns, pricing nobody has announced and, for agency teams, anything a client has shared in confidence. Strip names and details out of reviews and survey comments before you paste them, because the themes are what you need, not the people.

For the team itself, the rule is short. Marketing work happens in the tool your business has approved, signed in with a work account, because that's the account whose data terms your business has agreed and can manage. Our article on whether it's safe for staff to use ChatGPT and Copilot goes further on policy.

Which tool should a marketing team use?

The one your organisation has approved, used the same way by everyone. If your business runs on Microsoft 365, Copilot works inside Word, Outlook, PowerPoint and Teams, where your briefs, decks and meetings already live, and our Microsoft Copilot training works through those apps. If you've approved ChatGPT or Claude, both handle research, long drafts and repurposing well once they're briefed properly.

What keeps ten people sounding like one brand isn't the tool. It's a shared voice brief, a shared set of prompts and one sign-off process.

How do we get the whole team doing this?

Train them on their own work. Our AI training for marketing teams builds every exercise from your briefs, your copy and your channels: writing the voice brief, setting up projects, drafting and repurposing, and checking claims before anything is published. It runs privately for groups of up to 12, at your premises anywhere in the UK, at a venue or live online, with one price per group confirmed before you book.

Attendees rate our workshops 5/5, and responsible use runs through every session. DLR Media has signed the Responsible AI Use Campaign's 10 principles, which is why accuracy and human sign-off are built into each exercise. If other departments need training as well, our AI training for staff page sets out the options.

Frequently asked questions

Can AI write in our brand voice?

Yes, when it's briefed properly. Give it a voice brief and real examples of your best work, stored in a project, and first drafts need editing rather than rewriting.

Do we have to tell people an ad was made with AI?

No, not as a blanket rule. The ASA says there's no blanket legal requirement in the UK to disclose AI use in ads, but an ad that misleads breaks the rules whether AI made it or not.

Should we use AI to write blog posts?

Yes, for research, outlines and first drafts. A person who knows the subject edits every post, adds what only your business knows and checks every fact before it's published.

Which AI tools do you cover in marketing training?

ChatGPT, Claude and Microsoft Copilot as standard, plus any other AI tools your business has approved.

Plan your marketing team's AI training

Tell us which AI tools you're licensed for and which channels your team writes for, and we'll build the exercises around your own campaigns. Call +44 1382 699595 or book marketing team training.