New Brunswick legislator inadvertently reads AI prompt during floor speech
The Progressive Conservative member of the New Brunswick legislative assembly drew mainstream media attention after reciting an apparent large language model instruction during a parliamentary debate.

Bill Oliver, a Progressive Conservative member of the New Brunswick legislative assembly, has drawn widespread attention in Canada after reading an apparent large language model prompt instruction aloud during a floor speech. The incident, which occurred last month, has since gained traction on social media platforms including Reddit and Threads, prompting coverage from major national outlets such as the Canadian Broadcasting Corporation and The Toronto Star.
During the debate, Oliver addressed concerns regarding advocacy offices, noting that citizens often develop expectations that exceed the powers granted to such entities. He then proceeded to read the following line verbatim: “here’s a more natural, flowing version of that section that reads like a legislative speech rather than a series of short points.” The phrasing bears the distinct hallmarks of an AI model presenting an alternative stylistic option in response to a user prompt.
The error went largely unnoticed at the time of delivery but has since become a focal point for debate regarding the integration of artificial intelligence in political communication. The Toronto Star characterised the incident as indicative of a broader societal fracture, describing it as a sign of a divide between elites who are “only too happy to delegate their duties to the Borg” and the public, who view such practices as objectionable.
This event aligns with a wider trend of professionals facing scrutiny when the use of generative AI is exposed in their work. Lawyers, authors, journalists, and academics have previously encountered similar embarrassment when hallucinated errors or prompt instructions were revealed in ostensibly human-written documents. The incident underscores the risks associated with relying on automated tools for high-stakes public communications.
A Duke University study noted last year that workers often conceal their use of AI due to perceptions from colleagues that such reliance indicates laziness or replaceability. While politicians have long relied on speechwriters, Oliver’s apparent failure to remove an AI instruction from his script represents a distinct challenge in the era of automated content generation. It remains unclear whether Oliver was aware he was reading the prompt or the extent to which his speechwriting process relied on such technology.

