AI Agents as Users

Summary: AI agents now interact with digital interfaces alongside humans. Designing for both requires rethinking what "user" means and prioritizing accessibility.

The design community has spent decades refining what it means to design for users. We study their behaviors, map their journeys, test our assumptions against their needs. Our discipline is, and should be, oriented around understanding who is using the product or service, and designing accordingly.

AI agents (systems that pursue a goal by iteratively taking actions, evaluating progress, and deciding its own next steps) are now interacting with the same digital interfaces we design for people.

They navigate websites, fill out forms, compare options, and execute transactions. They do this crudely, often unreliably, and with significant limitations. In functional terms, they are users of our interfaces, even if we haven't recognized them as such. There's a conceptual shift required to account for this reality. A core assumption needs updating: "user" is no longer synonymous with "human."