OPINION: AI Shortcuts in Mayoral Race Demand Immediate Transparency
AI-generated image · Bay Street Wire
As candidates deploy algorithms to engage voters, City Hall risks a crisis of credibility without strict rules to prevent the illusion of consensus.
The race for Toronto's mayor is rapidly evolving into a testing ground for artificial intelligence, and the lack of transparency is alarming. As first reported by CBC Toronto, candidates are framing these tools as mere efficiency boosters, but the potential for algorithmic manipulation of the public discourse is a red flag for any voter who values authenticity.
Recent reporting from CBC Toronto highlights a growing divide over how AI is being integrated into campaign strategies. Coun. Brad Bradford has faced criticism following a National Observer report alleging his campaign utilizes an AI app to provide supporters with online talking points. The report further alleged the app rewards users for sharing phone contacts and posting social media comments.
Bradford has denied that users receive rewards, asserting that the app is simply a tool for connecting with voters. Laryssa Waler, a spokesperson for Bradford, told CBC Toronto that suggested content is created by volunteers rather than AI, comparing the practice to distributing traditional lawn signs and literature, only faster. Waler further claimed that every comment is reviewed and posted by a real person using their own account.
However, the Bradford campaign has pointed to Mayor Olivia Chow's use of Nationbuilder—organizing software that Bradford's team claims contains built-in AI elements. In response, Zeus Eden, a spokesperson for Mayor Chow, told CBC Toronto that her team does not use AI and instead employs writers, videographers, and designers. Eden stated that the campaign has disabled AI features on any software that utilizes them and does not use apps that require the surrender of personal information.
Other candidates are taking a middle path. Chris Alexander told CBC Toronto that his campaign uses AI for spelling, grammar, and some research, though he emphasizes that sources are verified before conclusions are drawn. Alexander warned that trust in City Hall is already thin and argued that voters should not receive information dependent on machines and algorithms.
The danger here isn't just about efficiency; it's about the distortion of democratic debate. Anatoliy Gruzd, a professor of information technology management at Toronto Metropolitan University and co-director of the TMU Social Media Lab, told CBC Toronto that AI can be used to organize social media "mobs" to post specific messages. According to Gruzd, this creates a false "illusion of consensus" on issues where a majority may actually disagree.
Adding to this risk is the inherent inaccuracy of the technology. Aengus Bridgman, a public policy professor at McGill University and director at McGill’s Media Ecosystem Observatory, told CBC Toronto that AI models are often misleading and produce inaccurate content that carries an "undeserved" veneer of credibility.
As Alex Marland, a politics professor at Acadia University, noted to CBC Toronto, the use of AI is becoming a thorny issue. If voters cannot trust the technology, they may struggle to trust the candidates who employ it. Without strict transparency rules, we are leaving the door open for deep-fake promises and manufactured support to steer the future of our city.