Hi everyone! Speech recognition is not currently one of the technology categories for which we at Instructure have an officially supported/recommended solution. However, we strive to engineer Canvas to web and accessibly standards which should allow standards-based technologies that are widely compatible on the web to work properly with Canvas.
The World Wide Web Consortium (W3C), which is responsible for the Web Content Accessibility Guidelines (WCAG), has a web page about Speech recognition and accessibility: https://www.w3.org/WAI/perspective-videos/voice/
As that page discusses, keyboard compatibility and proper labels/identifiers are key mechanisms for supporting speech control. Canvas is developed and tested to WCAG standards in these areas. You can read more about these standards in the Canvas VPAT: https://www.instructure.com/canvas/accessibility