AI in Political Campaigns: Trust and Transparency at Stake
AI tools in political campaigns raise authenticity concerns as Toronto mayoral candidates debate their use. Learn what experts say about voter trust, algorithmic bias, and campaign technology.
When Campaign Apps Go AI: The Toronto Mayoral Race Controversy
Political campaigns are quietly adopting artificial intelligence tools to amplify voter outreach, but a recent disclosure in Toronto’s 2026 mayoral race has exposed a tension that’s unlikely to disappear: voters want to know if they’re engaging with humans or machines. Councillor Brad Bradford’s campaign came under fire after using an AI app to supply talking points to supporters, sparking a broader debate about authenticity in digital-age politics.
The controversy isn’t really about whether AI belongs in campaigns—it’s about transparency, trust, and what voters actually want to see from the candidates they elect. As more municipal campaigns embrace software that generates content suggestions and analyzes voter data, the question becomes: how much AI is too much?
What Bradford’s Campaign Actually Did With AI
Bradford’s team deployed a campaign app that used artificial intelligence to suggest online talking points for supporters to share across social media. According to reports, the app also allegedly offered rewards for users who posted messages and shared their phone contacts with the campaign.
When confronted, Bradford denied the rewards claim, stating his team was simply “connecting with voters” through technology similar to what other campaigns use. His campaign spokesperson clarified that suggested comments and messaging are “reviewed, edited and always posted by a real person—the supporter, from their own account, under their own name.” The campaign characterized the practice as equivalent to distributing traditional literature and lawn signs, just faster.
However, the statement revealed a key detail: suggested content was created by volunteers on the team, not by AI itself. This distinction matters because it suggests the AI was being used more for distribution and amplification than for actual content creation.
How Competing Campaigns Are Handling AI Tools
The Bradford controversy prompted other mayoral candidates to disclose their own relationships with artificial intelligence. Olivia Chow’s campaign said it does not use apps requesting personal information and has deactivated AI features on any software that includes them. Instead, Chow’s team hires human designers, videographers, and writers to create campaign content directly.
Chris Alexander’s campaign acknowledged using AI for research purposes but emphasizes verifying all sources before drawing conclusions. His team also uses the technology for basic functions like spell-checking and grammar review.
Bradford’s team pointed out that Chow’s campaign uses Nationbuilder, organizing software that contains built-in AI elements. This counterpoint highlights a murky reality: AI is becoming so embedded in standard campaign technology that complete avoidance may be unrealistic.
Why Experts Worry About AI-Generated Consensus
Anatoliy Gruzd, a professor of information technology management at Toronto Metropolitan University and co-director of the Social Media Lab, identifies a specific risk: AI-powered apps can manufacture the appearance of widespread support when only a fraction of voters actually hold that view.
When a campaign app coordinates dozens or hundreds of users to post the same AI-suggested message, it creates what Gruzd calls “an illusion of consensus on a particular issue when the consensus may not exist.” On social media, this artificial amplification can drown out genuine grassroots discussion and mislead voters about true public opinion.
The concern extends beyond mere manipulation. McGill University public policy professor Aengus Bridgman notes that AI models frequently produce inaccurate content but are presented with unwarranted credibility. When voters encounter AI-generated talking points without knowing their origin, they may attribute more authority to them than they deserve.
The Authenticity Problem: Can Voters Trust Tech-Heavy Campaigns?
Alex Marland, a politics professor at Acadia University who researches campaign strategy, identifies an emerging voter concern: if people don’t trust artificial intelligence, they’ll question whether they can trust a candidate using it. Authenticity has always mattered in politics, but in an AI era, it becomes inseparable from transparency about which tools campaigns deploy.
Marland draws a distinction between ethical and unethical AI use. Deploying AI to help under-resourced campaigns compete is fundamentally different from using it to attack opponents or generate misinformation. The technology itself isn’t inherently problematic—the application is.
This mirrors historical precedent. When television advertising first entered political campaigns, fears circulated about the “hypodermic needle effect,” a belief that broadcast messages could inject ideas directly into voters’ brains. Over time, voters became media literate. Marland suggests the same learning curve will apply to AI, but only if campaigns operate transparently.
Can AI Ever Replace Doorstep Campaigning?
Brett Thalmann, a political organizer and founder of the Next Campaign Summit—an annual Toronto conference focused on campaign technology—sees AI as a tool for scaling limited resources. Campaigns use AI for content generation, data analysis, and voter identification, allowing smaller teams to reach more people.
Yet Thalmann emphasizes that “authenticity in politics is critical,” and campaigns must be careful about using AI to generate content that voters see directly. When an algorithm writes the message voters read, the human connection breaks down.
Marland agrees unequivocally: no technology, AI or otherwise, replaces direct voter contact. Winning campaigns still require knocking on doors, making phone calls, and having face-to-face conversations. AI can amplify reach and assist with data processing, but it cannot substitute for the human touch that persuades voters on a personal level.
What This Means for Future Municipal Campaigns
As AI becomes cheaper and more accessible, municipal campaigns—often run on shoestring budgets—will face pressure to adopt these tools. Smaller candidates competing against well-funded incumbents may view AI-powered apps as a necessary equalizer.
The challenge is maintaining voter trust while leveling the playing field. Campaigns that disclose their AI use, limit it to back-office functions like data analysis, and keep human judgment at the center of voter communications are more likely to build credibility.
Those that use AI to generate misleading talking points, manufacture false consensus, or deceive voters about the source of messages risk eroding the already fragile trust between politicians and the public.
Common Questions About Campaign AI Tools
Are campaign apps that use AI legal? Yes, using AI in campaigns is legal in Canada, but campaigns may face scrutiny if they use AI to spread false information or operate deceptively without disclosing how technology shapes voter communications.
How can I tell if a campaign message came from AI? Currently, there’s no reliable way for individual voters to identify AI-generated political messages. This is why expert recommend voters research candidates’ technology policies directly and ask campaigns how they use AI before supporting them.
Are all campaign apps using AI now? No. Some campaigns avoid AI entirely or use it only for backend data work. Others, like Bradford’s, integrate it into voter-facing features. It depends on the campaign’s resources, philosophy, and transparency approach.
The Toronto mayoral race has made clear that the conversation around AI in politics isn’t going away. Voters should expect more candidates to address their AI use head-on in future elections, particularly at the municipal level where resource constraints make technology adoption tempting. The key is ensuring that transparency, not deception, guides how campaigns embrace these tools.
