AI-assisted email copy + design: A designer's workflow in 2026

Let me be upfront about something: I don't use AI to replace what I do. I use it to get to the good part faster.

A year ago, the conversation around AI and design was mostly theoretical: a lot of "what if" and not enough "here's what I actually did this morning." In 2026, that's changed. There are real workflows now. Real time savings. And, honestly, some moments that feel a little bit like design superpowers.

This is my actual workflow for using AI to support email copy and design at Scalero. It's not a tutorial on a specific tool. It's more like a behind-the-scenes look at how this stuff fits into a professional design practice.

The shift: from blank canvas to conversation

The hardest part of any creative project isn't the execution. It's the first 20 minutes, when you're staring at a blank Figma frame or an empty subject line field and trying to get something worth reacting to on the screen.

AI has mostly solved that problem for me. Not because it always gives me the right answer, but because it gives me something to push against. I'll describe a campaign goal to Claude, the audience, the product moment, the tone I'm going for and ask it to generate three or four directions for the headline and hook. I rarely use any of them verbatim. But they almost always unlock the angle I actually want to pursue.

That's the real value. Not automation. Acceleration.

How I use AI for email copy

Starting with direction, not drafts

My first AI prompt in a campaign is rarely "write me an email." It's more like: "Here's the offer, here's the audience segment, here's what the last email in this flow said. Give me five different emotional angles I could lead with."

That kind of prompt gives me a strategic overview I can evaluate before I've written a single word. It surfaces angles I might not have thought of, and it forces me to articulate the brief clearly which is useful regardless of whether I use any of the output,

Iterating on copy, not starting from scratch

Once I have a direction, I'll draft a headline or body copy section myself, then ask AI to make it shorter, punchier, or more conversational. Back and forth, back and forth. It's like having a copywriter in the room who has infinite patience for revision and never takes feedback personally.

The key is staying in the driver's seat. I make the calls. The AI handles the legwork.

Subject line generation at scale

For subject lines specifically, AI is genuinely great at volume. I'll feed it the email content and ask for 15 variations across different tones and angles: urgency, curiosity, social proof, benefit-led. Then I narrow it down to three or four worth testing. We actually built a free AI subject line generator for exactly this use case.

How I use AI for design

Layout exploration without the Figma time

The Figma and Anthropic integration has been a game-changer for rapid exploration. I covered this in depth in my earlier post on how Figma and Claude are changing the way designers create, but the short version is: I can now go from a rough concept to a working layout prototype in a fraction of the time it used to take.

For email specifically, I use this to explore layout variations for campaigns, different ways to structure a hero, different approaches to a multi-product promo, different visual hierarchies for a plain-text-style vs rich-media send.

Writing copy and design in parallel

Here's something that's changed about my process: I no longer treat copy and design as sequential steps. I do them together.

I'll have a rough layout in Figma on one screen and a Claude conversation on the other. As I move modules around, I'm also refining the copy that lives inside them. The two inform each other in real time. A layout change might suggest a shorter headline. A copy direction might call for a different visual hierarchy.

This parallel workflow has cut our concept-to-review time significantly on campaigns. And it produces better work, because design and copy aren't fighting each other at the end they were built together.

What AI still can't do (and shouldn't)

AI doesn't know your client's brand the way you do after six months of working with them. It doesn't know that the last campaign with an urgency-heavy subject line underperformed, or that a particular segment responds better to long-form storytelling than punchy bullets.

Context is still yours to bring. Taste is still yours. The final judgment call, “does this actually feel right?,” is still a human one.

The teams getting the most out of AI right now aren't the ones using it to replace creative judgment. They're the ones using it to protect time for creative judgment. That was one of the themes we explored in the 2026 lifecycle marketing trends, and I think it holds.

A sample workflow, start to finish

• Brief the AI: share campaign goal, audience segment, tone, constraints

• Generate strategic directions: 4–5 angles to evaluate before writing

• Draft in parallel: rough layout in Figma, copy iterations in Claude

• Refine: tighten copy, test layout variations, align the two

• Review with team: share the Figma frame + copy doc together, not separately

• QA + ship: dark mode check, mobile check, final copy read

AI in a design workflow isn't about working less. It's about working on the things that matter more. The creative decisions, the strategic calls, the nuances that only someone who actually knows the brand and the audience can make.