Canadian lawmaker Bill Oliver sparked widespread online debate after accidentally reading an AI chatbot’s writing prompt aloud during a session of the Legislative Assembly of New Brunswick, reading transition phrases meant for a chat interface straight into the official record on July 9.
Legislative Assembly Gaffe Goes Viral
A routine debate in the Legislative Assembly of New Brunswick took an unexpected turn when a veteran politician delivered an AI-generated prompt directly into the microphone. Bill Oliver, a member of the Progressive Conservative Party, was addressing the chamber regarding the Office of the New Brunswick Advocate when the slip occurred.
Oliver began his address with standard legislative rhetoric before stumbling into text that clearly originated from a large language model. Video footage captured the lawmaker declaring that Public confidence in the office of an advocate matters,
before immediately reading a transitional cue generated by artificial intelligence.
Instead of pausing or correcting the error, Oliver remained oblivious and continued reading the newly styled text aloud. According to reports from The Independent, the phrase matched typical output patterns produced by AI systems when asked to rewrite bullet points into prose.
Public Backlash Over Political Preparation
While the video clip initially circulated locally on the New Brunswick subreddit, it quickly spread across platforms like X and Reddit. Mainstream outlets across Canada and internationally picked up the story, drawing sharp criticism from social media users who questioned the preparation habits of elected officials.
Commentators argued that the primary failure was not the use of artificial intelligence itself, but rather a total lack of oversight. Critics noted that whoever compiled the speech—whether Oliver or an assistant—failed to review the final document before it reached the legislative podium.
Comparisons to Legal Sanctions for AI Errors
The incident triggered comparisons to professional fields where using artificial intelligence without verification carries severe penalties. Observers online pointed out that lawyers who submit unverified AI-generated text in court filings face sanctions from judges.
Recent legal precedents underscore the risks of unvetted AI output in professional settings. Just last month, a US appeals court sanctioned two lawyers for filing documents with AI words. During the exact same week, a Mississippi lawsuit resulted in penalties for four attorneys on both sides of the case, with all four fined and two barred from appearing at the US district court for the next two years.
Response From Oliver’s Office
As the video accumulated views and drew widespread media coverage, Oliver’s office declined to comment when contacted by news organizations. Representatives did not clarify whether only the specific transitional phrase was generated by an AI tool or if the entire address was drafted using an automated writing assistant.
Neither legislative leaders nor Oliver have issued a formal apology or explanation regarding the floor speech. The incident leaves open questions regarding whether legislative bodies will implement formal guidelines or restrictions governing the use of generative AI tools in drafting official speeches and documents.
Worth a look
