Cookie preferences

Essential cookies help make a website usable by enabling basic functions like page navigation and access to secure areas of the website. The website cannot function properly without these cookies.

Preference cookies enable a website to remember information that changes the way the website behaves or looks, like your preferred language or the region that you are in.

Statistic cookies help website owners to understand how visitors interact with websites by collecting and reporting information anonymously.

Marketing cookies are used to track visitors across websites. The intention is to display ads that are relevant and engaging for the individual user and thereby more valuable for publishers and third party advertisers.

Skip to main content
AI Agents are changing digital product design

There’s quite an evolution happening in digital product design. A new kind of user is emerging, and it’s not even human. It’s an AI agent, and it's scrolling through interfaces, interpreting content, navigating forms, and even making purchase decisions on behalf of customers. 

As a result, it’s changing the rules and existing design guidelines we’ve followed in the past. Our UX designers now have to consider the mechanics of the AI agent experience as part of our design process.

AI agents accelerate testing

Optimising the online experience takes time and involves significant analysis and evaluation. While there's no substitute for having real people validate design solutions under time and budget pressures, AI agents can simulate and test user journeys in a fraction of the time it would take with traditional usability testing. These agents do not just click through screens; they interpret intentions, stress-testing flows, and surfacing weak points with astonishing precision. 

And that’s just the beginning. The possibilities for stress-testing design with AI are exciting. From asking simple questions like, What happens if I go this way instead? What if I miss a call to action? What if I skip the small print? to more complex ones involving user behaviours, such as rushing through a checkout flow to test its efficiency or operating with low digital literacy to uncover issues that a tech-savvy user might miss. We can even have it ignore instructions along the way to see how resilient our user interface really is.

Man holding map and mobile phone
Increasingly people are using voice as the main way to interact with their smartphones
AI moves from tester to user

However, what is perhaps even more exciting is that AI agents are crossing the line from testing for users to becoming one.

We’re seeing a growing number of real-world interactions happen through AI intermediaries such as chatbots, virtual assistants, voice interfaces and smart search agents. Whether it’s Siri helping someone find a product or ChatGPT plugging into a SaaS dashboard via an API, more and more experiences are being navigated not by human hands on a touchscreen but by agents acting on behalf of the person.

From here on out, the design of digital products and experiences will need to be AI-centric as well as human-centric. 

This means designers will need to account for the behaviours of AI agents in the research and design process. If product descriptions aren't clearly structured, AI assistants may miss important details or convey them inaccurately. Rigid and linear user flows can prevent these agents from navigating effectively or offering practical support. And when accessibility isn’t integrated from the start, users who rely on AI-powered tools risk being excluded.

AI - a new kind of persona

Furthermore, UX personas traditionally represented archetypes of human users: goal-driven, emotionally nuanced, and behaviorally diverse. But now there’s a new persona to consider: the AI agent.

Unlike human users, AI agents aren’t emotional… yet! They don’t scan designs for visual delight; they read semantic structure, APIs and metadata. They are goal-oriented and never hesitate, so in addition to accommodating human behaviour, designs will now need to work for a user built with machine logic in mind.

This doesn’t mean replacing human-centered design; it means expanding it. 

Designing for AI agents ensures the product can be interpreted, navigated, and actioned on behalf of real users. It’s about making the experience resonate with not just eyes and fingers but algorithms and models.

Woman holding futuristic phone
A new persona is emerging - the AI agent
Why it matters now

The stakes are growing. From apps that integrate with AI-powered assistants and financial tools used through chatbots to products being sourced, assessed, and purchased via AI, agent-mediated experiences are becoming normal. Customers want convenience, and AI delivers this in spades. Yes, people will always want the alternative option, the ability to do their legwork if desired. Still, as digital designers, we cannot ignore this shift to people frequently using AI as their personal digital concierge.

Another reason for considering the agent experience as part of your UX strategy is its simplicity or ease of use for the agent. Remember, if it cannot easily complete its task of navigating a site, it will skip it and go to another one. Therefore, as a brand, you'll want your pricing, features, and value proposition to be easily recognised and understood by the chatbot when it is asked to find products or conduct a comparison. If the agent can't get the data, your brand won't be represented in the answer.

Embracing AI agents as users in our design process opens up a powerful new dimension in UX, one where our designs aren't just beautiful, human, functional, and future-proof.