AI Misstep in Canadian Legislature
Canadian Members of Parliament recently showcased a somewhat awkward use of AI when one legislator inadvertently read out formatting instructions meant for an AI chatbot during an official speech.
Bill Oliver, a member of the Progressive Conservative party in New Brunswick, unintentionally revealed AI assistance while discussing the complexities of creating an advocacy office. During his speech, he stated, “One of the dangers of establishing an advocacy office is that citizens often have expectations that exceed the powers actually vested in the office.” However, he then proceeded to read phrases that seemed to be AI-generated, saying, “Here’s a more natural, flowing version of that section that reads more like a speech to Congress than a series of short points.” This type of phrasing is typically associated with AI chatbots that aim to offer different writing styles based on user inputs.
This unusual moment somewhat slipped under the radar at first but caught fire on social media platforms like Reddit and Threads, garnering significant attention. Major news outlets across Canada, including the Canadian Broadcasting Corporation, reported on the incident.
A commentator referred to the blunder as indicative of “the growing divide in our society between the elites who are happy to hand over their duties to the Borg and the masses who are uncomfortable with it.”
While it’s not uncommon for politicians to read speeches penned by others, and this isn’t the first time AI has been involved in such processes, Oliver’s reading of apparent AI instructions without recognizing it in real-time highlights a new level of embarrassing exposure for AI use.
Moreover, incidents aren’t limited to Canadian politics. Reports have surfaced of law firms, like Sullivan & Cromwell, facing backlash for submitting court documents rich with inaccuracies resulting from AI misinterpretations.
A senior partner at Sullivan & Cromwell recently acknowledged to Manhattan Chief Justice Martin Glenn that previous legal filings included incorrect quotes and elements they termed as AI hallucinations. According to Andrew Diedrich, co-head of global finance and restructuring at the firm, these “hallucinations” refer to when AI tools concoct false citations, misquote sources, or create legal references that do not exist. He expressed regret for these occurrences.
As both politicians and legal entities grapple with the implications of AI, there’s a growing realization of both the vast potential and the inherent risks it presents to society. A new book titled Code Red: Left, Right, China, and the Race to Control AI aims to provide insights on how movements can position themselves regarding AI challenges without ceding control to dominant tech interests or external powers.
