Email in the Age of AI

Email blog post

Good email design has always been about maximum accessibility, and now that AI is taking over the inbox, accessibility has become important for a whole new reason.

The Case for Accessibility

Over the years, I’ve seen accessibility conversations move to the forefront, with best practices baked into design processes. After all, as designers of products and experiences, we want to connect with all people, not just some. On the web, making experiences both accessible and visually engaging is relatively straightforward. In email, however, where speed and aesthetics have long been the primary focus of design, achieving both is a real design challenge. In the past, email designers often tested how much they could hack the platform to create visually stunning pieces that would surely move the needle and compel clicks.

That framing is outdated. And I would argue it isn’t good design.

It’s important to remember that as designers working in marketing, we’re in the business of communication. And marketing communication is only effective if it reaches as many people as possible. If you believe that, then accessibility has to be about more than compliance, it has to be about intelligibility.

According to the World Health Organization, more than one billion people globally live with a disability at the time of this writing. Roughly 285 million people are visually impaired and rely on assistive technologies like screen readers to access digital content. Accessible email design helps ensure:

  • Logical reading order
  • Clear hierarchy
  • Meaning that doesn’t rely solely on images

This is table stakes for inclusive communication. If your message can’t be understood, it doesn’t matter how good it looks.

But there’s another audience reading your emails now, and it’s making design rigor even more important.

Your Emails Aren’t Just Read by Humans Anymore

Beyond assistive technologies, emails are increasingly being analyzed and summarized by AI-powered systems. Tools from Apple (Apple Intelligence), Google (Gemini), and Microsoft (Copilot) analyze content to surface summaries, highlights, or priority information for users. This changes how emails are interpreted, at least on the client side. AI systems don’t ‘see’ your email the way a sighted person does—generally they operate on machine-readable inputs: text order, structure, headings, lists, links, and labels. This is largely the same underlying layer that screen readers rely on. Now, AI is also looking at behavior patterns in the inbox as well, but from a creative perspective, we have direct influence on optimizing the email itself for the inbox. Emails built with semantic structure and clear hierarchy are less ambiguous to machines and more useful to all humans, regardless of ability. Headings signal importance. Lists convey grouping. Explicit CTAs describe intent. Plain, scannable language reduces ambiguity.

In that sense, accessibility practices reduce the risk of misinterpretation by both assistive technologies and AI systems.

What AI Is Doing in the Inbox

A study by the Nielsen Norman Group found that users depended heavily on the friendly from (the sender name that shows in the inbox), subject lines and preview text together to determine if the message was relevant enough to open. That layer still matters, but it’s no longer the only one.

Some inbox experiences now generate AI-written summaries of emails. Apple Intelligence, for example, can replace traditional preview text with a system-generated summary based on the email’s content and structure. This varies by device, OS version, and user settings, but the direction is clear: inboxes are becoming more interpretive, not just presentational.

It’s important to note that these new systems are designed to help users quickly understand what an email is about—which is a good thing. But they’re not relying on aesthetics to do so.

Rather than trying to game or “hack” these systems, accessibility best practices align with how they already work. Clear structure, explicit language, and meaningful hierarchy give both humans and machines better inputs.

An accessible email may not guarantee better performance, but it positions your message to be understood as intended, even when the inbox itself is acting as a gatekeeper.

Start With the Basics

Our first impulse as designers is often to lean into highly creative, visually rich layouts. But image-heavy emails collapse the moment images fail to load, or when a screen reader can’t interpret them.

The basics:

  • Ensure the message works without images
  • Use clear hierarchy and meaningful structure
  • Write CTAs that make sense on their own

A simple gut check:

  • Can the email be understood without images?
  • Can the message be summarized from the structure alone?
  • Does the CTA make sense on its own?

If you can’t answer “yes” to these, you risk misinterpretation—whether the reader is human, assisted, or AI.

What About Visual Identity?

I’ve spent the last few paragraphs intentionally de-emphasizing the visual component of email. But the fact remains: as designers, we can’t only focus on intelligibility. I’m not saying visuals don’t matter—I’m saying we need to start somewhere else. There is a balance of utility and delight that we’re trying to achieve within strict constraints, and the right combination of email best practices, on-brand copywriting, and strategic application of visuals, helps to deliver the spirit of the brand to the reader in an inclusive way.

The real work then becomes delivering on the brand promise once someone opens the email, and again when they click through to your brand’s content. Creating a meaningful, aesthetic experience that lives within the constraints of email is the gig.

Designing for the New Inbox

Automation will always be central to email marketing, and AI is making it more powerful than ever. As designers when we think of AI, our first question shouldn’t be how we can incorporate it into our workflow—that will come naturally as all new technologies do. The real puzzle will be how to design the next phase of human-computer interactions. We’re designing for clarity. We’re designing for inclusivity. Nail that, and the rules don’t limit you—they set you up for the inbox of tomorrow.

Share

Partner
with
us

Have a project in mind? Let's hear about it.