Your next user isn't human.
This week at Techstars Tech Central Sydney Accelerator, I heard a pitch from Sedso that stuck with me: brands have zero clarity on how AI perceives them.
When someone asks ChatGPT "what's the best Christmas gift?" what does it say? Why? Can you influence it?
Here's the shift we're missing:
We've spent decades optimizing for human users (curious, distractible, emotional). We charmed them with microcopy, guided them with visual hierarchy, converted them with psychology. AI agents don't work that way.
They show up with a mission: "book X", "compare Y", "find Z under $50." No wandering. No impulse clicks. No being charmed by "Grab your slice of sunshine" CTAs.
The fundamental shift? The agent isn't the interface. It's the user. And that changes how we design.
Instead of clever copy, we need clear pathways.
Instead of brand voice, literal labels.
Instead of visual storytelling, structured data.
"Grab your slice of sunshine" means nothing to an LLM. "Order lemon cake $12" does.
Here's where it gets interesting:
If you've designed for accessibility, you're already halfway there.
Screen readers and AI agents need the same thing: structured clarity, semantic HTML, descriptive alt text, logical hierarchy.
Accessibility isn't just ethics anymore. It's infrastructure.
This doesn't mean abandoning human-centered design. It means expanding who "user" includes.
The journey now looks like: Human → AI Agent → Service.
You're designing for both, the human who asks and the agent that executes.
Two things you can do now:
Audit for accessibility — If a screen reader can navigate your site, an AI agent can too. Check your semantic HTML, alt text, ARIA labels.
Track agent traffic differently — Agent behavior ≠ human behavior. Different users need different metrics.
The designers who figure this out first won't replace human-centered design. They'll just know what "human-centered" means now, designing systems that work for humans, executed by agents, at scale.
Sedso is onto something. The question is: are we paying attention?

