AI and Copywriters: What's Actually Happening and What to Do

The honest assessment

Copywriting was one of the first professions to get hit by generative AI, and it's been hit hard. I'm not going to dress this up. ChatGPT launched in November 2022, and by February 2023, freelance copywriting rates had already started falling. Content mills that used to pay humans £0.05 per word now pay AI £0.00 per word. The bottom of the market didn't just get disrupted. It evaporated.

Here's what AI can do right now. Write blog posts, product descriptions, email sequences, social media copy, landing page text, ad copy variations, SEO content, press releases, and marketing collateral. It can match brand voice guidelines if you give it examples. It can write in any tone, for any audience, at any length, in any language. It does it in seconds. And the quality is... fine. Not great. Not terrible. Just... fine. The kind of fine that's good enough for a lot of clients who were never paying for great in the first place.

What AI can't do is have a point of view. It can write competent sentences about any topic, but they read like someone who's never actually experienced anything. There's no scar tissue in the writing. No "I once worked with a client who..." No real opinion. No risk. It produces what i call "corporate smooth"... text that says everything and nothing simultaneously. You've read it. You've probably written it too, on a bad day. The difference is that a good copywriter can look at that smooth output and know it needs grit. AI doesn't know what grit is.

The area changing fastest is volume content. SEO articles, product descriptions for e-commerce sites with thousands of SKUs, social media calendars, internal communications. Any writing that's primarily functional rather than creative is being automated rapidly. Some agencies have already shifted to a model where AI writes and humans edit. Which means they need fewer humans. Considerably fewer.

Your exposure level: High

High. Properly high. Copywriting is one of the most exposed professions to generative AI, and denying it won't help.

The reason is that writing is literally what large language models do. That's the whole thing. You can't be a professional writer and wave away a technology whose primary capability is writing. The economic impact is already visible. A University of Oxford study found that freelance writing jobs on platforms like Upwork and Fiverr dropped significantly after ChatGPT launched. Rates for routine content work have fallen. Some agencies have been candid about reducing their writing teams.

But here's the nuance. "Copywriter" covers an enormous range of work. There's the copywriter who writes 500-word blog posts about accounting software for SEO purposes. And there's the copywriter who names brands, writes campaigns that make people feel something, and can distil a complex value proposition into six words on a billboard. AI is coming for the first one. It's not even close to the second one.

The writers who are surviving and even thriving right now are the ones who've accepted reality and adapted. They use AI for research, first drafts, and variations. Then they add the thing AI can't: a voice. An angle. A sentence that makes someone stop scrolling. The job has changed from "writing" to "thinking and editing, with AI doing the typing." If you can make that shift, there's actually a case that good copywriters become more valuable, not less. A copywriter who produces ten excellent pieces a week instead of three is worth more per hour. If your output quality stays high while your speed triples, the maths works in your favour.

The 90-day action plan

1. This week: write against AI. Take your last three briefs. Give them to ChatGPT and Claude. Let them write the copy. Now compare the output to what you wrote. Where are you better? Where is it better? (Be honest about the second one.) The gaps you identify are your survival skills. Learn them by name.

2. Week two: build an AI-assisted workflow. For your next project, use AI for the parts that don't require your brain. Research. First drafts. Variation generation. Outline creation. Then add your value in editing, tone, structure, and strategic thinking. Time the whole process and compare it to your old workflow.

3. By day 30: develop your voice intentionally. The one thing AI can't replicate is your specific voice. Your perspective, your rhythm, your cultural references, your ability to write a sentence that makes someone snort coffee out their nose. If you haven't consciously developed and documented your voice, do it now. Write a "style guide for me." It sounds narcissistic. Do it anyway.

4. By day 45: learn to prompt like a writer. Most people prompt AI like they're filling out a form. A good copywriter should be able to prompt with nuance. Specify not just what to write but how to write it. The cadence. The emotional register. The level of formality. The sentence length variation. Become the person who gets genuinely good output from AI because you know how to direct it.

5. By day 60: move toward strategy. Clients don't just need someone to write words. They need someone to figure out what words to write and why. Messaging frameworks. Brand voice development. Content strategy. Campaign concepts. Move your offering upstream, where AI is weakest and the budgets are bigger.

6. By day 75: build social proof around AI expertise. Write a LinkedIn post or blog about your experience using AI in your writing process. Case studies. Real numbers. Honest assessments. The copywriters getting hired right now are the ones who can say "I know how to use these tools" not the ones pretending the tools don't exist.

7. By day 90: restructure your pricing. Stop charging by the word or by the hour if you still are. Charge by the deliverable, the project, or the outcome. If AI makes you three times faster, hourly billing punishes you for being efficient. Project-based pricing means you capture the value of speed without racing to the bottom.

AI tools you should be using this week

• ChatGPT for Work — The most versatile writing tool available. Use it for first drafts, brainstorming angles, generating variations, and repurposing content across formats. Give it detailed briefs and examples of your style. The output improves dramatically when you give it better input. Garbage in, garbage out still applies.

• Claude for Work — i prefer Claude for longer-form writing and strategic work. It handles nuance better and is less likely to produce that generic "corporate smooth" tone. Particularly good for brand voice work, long-form content, and anything that requires sustained logical argument. Paste in examples of writing you admire and ask it to identify the stylistic patterns.

• Grammarly AI — Goes well beyond spell-check now. Use it to review AI-generated drafts for tone consistency, readability, and brand voice alignment. Also useful as a sanity check on your own editing. Catches the subtle tells that flag content as AI-generated.

• Perplexity for Research — Faster and more reliable than manual research for most content work. Gives you cited sources, which matters for credibility and fact-checking. When a client brief says "write about AI trends in healthcare," Perplexity gets you up to speed in minutes rather than hours.

What to say in meetings

When a client suggests replacing you with ChatGPT: "That'll get you competent copy. It won't get you copy that converts. I tested AI on your last three briefs and the difference in [specific metric] was significant. Happy to show you." Always have the data. Opinions are easy to dismiss. Data isn't.

If an agency is discussing AI integration: "I'm already using AI in my workflow. It handles the research and first drafts. I handle the strategy and the voice. My output per week is up 40% with no quality drop. Here's what that looks like." Be the person who's already doing it.

When someone says "AI writes as well as humans now": "It writes as well as average humans on average days. That's exactly the point. If you want average, it's free. If you want something that actually works, you need someone who knows the difference."

If the worst happens

If you're made redundant from a copywriting role, your core skill... the ability to communicate clearly and persuasively... is valuable far beyond writing copy. Content strategy, UX writing, brand consulting, communications management, editorial direction, and marketing leadership all draw on the same fundamental abilities. The market for "people who can write" is shrinking. The market for "people who can think and then write" is not.

Adjacent moves to consider: content strategist, UX writer, brand strategist, communications director, editorial consultant, or AI prompt engineer (yes, really... companies are hiring people who can get good output from AI, and writers are better at this than most). The corporate communications market is also robust because internal comms requires understanding organisational politics and sensitive topics that AI handles poorly.

Here's the uncomfortable truth and the hopeful truth wrapped into one. The era of making a comfortable living writing mediocre content is over. It just is. But the era of making a better living writing excellent content, faster, with AI tools, is here. The question isn't whether you're a writer. It's whether you're a good enough writer that AI makes you more valuable rather than redundant. If the answer is yes, charge accordingly. If the answer is "I'm not sure," you've got your 90-day homework.