Voice control (speech input)
Operating an interface with spoken commands, usually by saying the visible name of a button, link, field, or menu item.
See it
What it is
Voice control lets someone operate the interface by speaking commands such as 'Click Save' or by calling out an on-screen number. Apple's Voice Control, Android Voice Access, and Dragon turn those commands into clicks, focus changes, text entry, and navigation without requiring a mouse or touch gesture.
Visible labels are the contract. If a button says 'Send invoice', its accessible name should contain those words so the command a person can see is also the command the software can target. Native controls and unique, specific labels make this work with little extra code.
The classic failure is replacing visible text with a different aria-label. A button that visibly says 'Search' but is named 'Find products' in the accessibility tree may ignore 'Click Search'. Icon-only controls also force users to display a numbered overlay, so give common actions visible text when the layout allows it.
Ask AI for it
Make this interface operable with Apple Voice Control. Ensure every native button, link, and form control has an accessible name that contains its visible label text, satisfying WCAG 2.2 Label in Name, and put those words at the start of the name where the wording allows. Remove aria-label values that replace visible wording with different phrases; use aria-labelledby when visible text should supply the name. Give repeated actions unique visible labels, add visible text to important icon-only controls, and keep target sizes usable. Then write a manual test using the Voice Control commands 'Show names', 'Click [visible label]', and 'Show numbers', listing the expected action for each control without inventing speech output.