Canadian Legislator Reads AI-Generated Prompt Verbatim in Parliament

Based on an article by Kyle Orland for Ars Technica.

A new controversy involving the use of AI in legislative proceedings has emerged in the New Brunswick Legislature in Canada. Bill Oliver, a member of the Progressive Conservative Party and a provincial legislator, committed a blunder last month when he read out what was clearly an AI-generated prompt verbatim during a parliamentary speech.

Initially, the incident attracted little attention. However, a video of the speech began circulating on social media platforms like Reddit and Threads earlier this week. The matter quickly gained traction, prompting coverage from major Canadian media outlets such as the Canadian Broadcasting Corporation (CBC) and the Toronto Star. The Toronto Star described the incident as indicative of a growing societal divide, where elites willingly delegate their responsibilities to collective intelligence while the general public resists such trends.

Sequence of Events

During his speech last month, Oliver stated, “One of the risks of establishing advocacy bodies is that citizens often have expectations that exceed the authority actually granted to these bodies.” He then added:

“Here’s a more natural, flowing version of that section that reads like a legislative speech rather than a series of short points.”

This phrasing is a textbook example of how large language models (LLMs) present alternative stylistic suggestions when responding to user prompts. It is believed that Oliver, unaware that this was an AI-generated instruction, read it aloud as part of the official parliamentary record.

Blurring the Lines Between Politics and AI

It is not uncommon for politicians to read pre-written scripts verbatim. Moreover, it’s not difficult to imagine that the use of LLMs to craft speeches is becoming increasingly common among legislators. However, the fact that a nonsensical, AI-generated sentence was read aloud in a public setting without being recognized underscores a troubling new dimension of AI use in politics.

This incident follows a series of similar cases where AI-generated content has caused problems in professions such as law, journalism, writing, and academia. In each instance, the issue arose because users failed to verify the AI’s output, which sometimes contained false or nonsensical information (known as “hallucinations”). A study conducted by Duke University last year revealed that workers often conceal their use of AI tools out of fear of being perceived as “lazy” or “replaceable” by their colleagues.

Editorial Opinion

The crux of this issue is not whether politicians should use AI, but rather the quality and transparency of its use. Parliamentary speeches are a cornerstone of democratic communication, and how these speeches are crafted is a matter of public concern.

Oliver’s misstep is a textbook example of how uncritical and unconscious use of AI tools can actually degrade the quality of work. Over the next three to six months, we are likely to see accelerated efforts to establish clear guidelines for AI use in provincial and federal legislatures.

From a long-term perspective, this case highlights the paradoxical problem of the “stigma of AI usage.” If the culture of concealing AI use in professional settings persists, similar incidents will likely recur. Instead, there should be a push for politicians to publicly disclose their use of AI and clearly define the scope of its application. Such “AI disclosure rules” should not merely serve as ethical guidelines but function as a new standard of accountability in democratic processes.

This incident also raises questions about how Japan’s national and local legislatures would respond if faced with similar scenarios.

References

• “Canadian legislator reads out apparent LLM response in floor speech”, by Kyle Orland — Ars Technica, 2026-07-24T21:25:15.000Z (CC BY-NC-ND)
• Source URL: https://arstechnica.com/ai/2026/07/canadian-legislator-reads-out-apparent-llm-response-in-floor-speech/