Can AI write good website copy?

This morning I settled at my desk with a coffee and opened an email with fresh content for a client's website. It soon dawned on me that I had read it somewhere before. Everywhere, actually.

Different businesses. Same words, same phrases, same rhythm. That pattern.

Finding the right words is harder than most people realise and far more important than most assume. I remember when getting content from clients was a long and painful process — challenging them to move past "A dynamic professional team" and "Delivering quality solutions."

Then AI arrived, and we thought the problem was solved.

Almost overnight, words became instant. Pages of content in seconds: business descriptions, social bios, website copy, email campaigns. Copywriters were the first profession declared "replaceable."

We were just swapping one problem for another.

From scarcity to sameness

Instead of too little copy, we now have too much of the same copy. The scale of it only becomes visible from where I sit — reading content across dozens of businesses and industries. What sounds fresh in isolation is painfully familiar in aggregate.

Millions of business owners are being served the exact same phrases by the same tool. What feels clever from an isolated perspective actively detracts from the brand when your audience is consuming multiple brand narratives at once.

LLM AI is exceptionally good at assembling sentences. What it cannot do is understand how those sentences actually sound — or what matters most for a specific business strategy. That's the risk.

This isn't an anti-AI argument. I use it myself: for brainstorming, refining language, checking grammar, exploring alternatives. Used well, it makes the process faster.

The danger is when creative thinking gets delegated to the machine entirely.

The phrases that follow you everywhere

AI is by nature parroting the most frequent language. Brand copywriting is about voicing differentiation. These two things are fundamentally opposed.

After reading hundreds of pages of client copy, certain phrases become almost comforting in their predictability. They travel from site to site, industry to industry.

Here are the ones I see most:

"We help businesses unlock their full potential"

"Delivering innovative solutions for modern businesses"

"Driven by innovation and powered by expertise"

"Helping brands grow, scale, and succeed"

"A forward-thinking company at the forefront of [insert industry]"

"Combining strategy, creativity, and technology"

"Built on trust, transparency, and results"

"Where innovation meets execution"

"Partnering with businesses to drive meaningful outcomes"

"Focused on delivering measurable outcomes"

"Helping companies navigate an ever-changing landscape"

"We pride ourselves on quality, reliability, and innovation"

"Transforming ideas into reality"

"Results-driven strategies that make a difference"

"Shaping the future of [insert industry]"

"Innovating today for a better tomorrow"

You've read all of these. So has everyone else.

How to tell if your copy has the same problem

We can teach AI to sound more like us — and it's worth doing. But the moment we stop challenging our own messaging, the brand loses its edge.

Run your copy through these seven questions before you publish:

1. Could this copy come from any company, or does it feel distinctly ours?

2. Does it communicate the tone, voice, and values of the brand?

3. Does every sentence serve a purpose, or is it filling space?

4. Is there one clear point a reader can take away?

5. Does it include real examples, proof points, or tangible detail?

6. Does it make the reader feel something, or just read something?

7. Is it aligned with the brand's positioning — not just the product?

If you're hesitating on more than two of those, the copy needs work.

What strong copy actually requires

Finding the right words is a creative process: brainstorming, stripping back the generic, digging for the thing that is actually true and specific about this business.

A few principles that guide how we approach it:

• Avoid being purely descriptive. Describe the experience, not just the feature.

• Add a point of view. Share an opinion and back it up with specific expertise.

• Every sentence should earn its place. Dead copy is worse than no copy.

• Support your message with strong visuals — copy and image should do different jobs.

• Less is more. Revisit your copy at least once a year. It gets stale faster than you think.

• Say something slightly uncomfortable. Safe messages are forgettable.

The thing AI can't fix

No brand can afford sameness. Words are abundant, but captivating stories are rare. Copywriting isn’t going anywhere. Humans still turn words into stories, and strategy still makes them matter.

AI raises the floor and lowers the bar at the same time. It means more copy exists, and less of it is worth reading. The businesses that understand this, and invest in the words that actually represent them, have a cleaner field to stand out in.

That's not a threat to copywriting. It's the argument for taking it seriously.