The structured semantic representation of a web page that assistive technologies and browser agents use to understand UI elements, roles, and state.
More about Accessibility Tree
The Accessibility Tree is a structured, semantic representation of a web page that browsers build alongside the DOM. It exposes each interactive element as a node with a role (button, link, textbox), a name (the label), a state (checked, expanded, disabled), and other properties. Screen readers and other assistive technologies use this tree to describe the page and let users navigate it.
The accessibility tree is also one of the primary inputs that browser agents use to "understand" a page. Where a raw DOM is noisy and full of presentational details, the accessibility tree is closer to the semantic shape of the UI: which things are clickable, what they are called, what state they are in. That makes it dramatically easier for an AI agent to decide what to do next.
For website owners, this means accessibility work pays a double dividend. Proper use of semantic HTML, ARIA roles and labels, and good focus management makes a site usable for people with disabilities AND makes it more reliable for AI browser agents trying to complete tasks. That overlap is one of the most underrated reasons to invest in accessibility for AI agent readiness.
