An MLA reads out an AI generated prompt on the floor of the legislature

“Here is a more natural and fluid version of this section, written as a parliamentary speech rather than a set of short talking points” — an insert from an AI text editing model that’s standard practice. It didn’t occur to anyone in the chamber, but the video has gone viral on Reddit and Threads and has been picked up by CBC and the Toronto Star.

Politicians using AI to create speeches isn’t anything new — but reading AI model prompts out loud without being aware it happened is a first in the long history of AI mistakes in public spaces. It brings us one step closer to asking the right question: not “should we ban AI?”, but “who is accountable for the content someone reads without reading it themselves?”