A growing share of Canadians view artificial intelligence as a threat rather than an opportunity, according to a new national survey from Research Co., with public unease about the technology intensifying over the past year.
The online poll, conducted among a representative sample of Canadian adults, found that three-in-five Canadians (60%) now believe AI is a threat to humanity, up 10 percentage points from a similar survey in July 2025. Meanwhile, the share who see AI as an opportunity fell to 31%, a six-point drop over the same period.
The generational divide on this question persists. Canadians aged 35 to 54 remain the least likely to view AI as a threat, at 44%, compared with 60% of those aged 18 to 34 and 63% of those 55 and older.
Calls to Slow Down or Halt AI Development
More than seven-in-ten Canadians say they want the development of AI either slowed down or stopped entirely. Just over half (51%) favour slowing the pace of development, while 18% believe it should be abandoned altogether — up from 13% who held that view in 2025. Only one-in-five Canadians (20%) think AI should continue to be developed as quickly as possible.
Engagement with AI-related news has also risen, with 63% of Canadians saying they followed AI stories closely over the past year, up seven points. That figure climbs to 70% among Canadians aged 18 to 34, 70% among Quebecers, and 74% among those in the highest household income bracket.
Widespread Worry Over Jobs, Data and Deepfakes
Concerns about AI’s societal effects remain high across the board. More than four-in-five Canadians say they are worried about AI contributing to less intelligent students in schools and universities (82%) and about AI displacing jobs currently done by humans (82%). Nearly three-in-four (74%) are concerned about AI causing an event that results in loss of human life.
Other worries cited by respondents include AI-generated deepfake videos confusing the public about political candidates’ positions (82%), deepfakes used to steal personal identity (81%), and the cost and energy demands of AI data centres (81%). Additional concerns include AI scraping websites and datasets for information and images (78%) and AI bots draining funds from bank accounts (74%).
About three-in-five Canadians say they sometimes feel it is hopeless to protect themselves from AI-related intrusions (61%) and expect AI to have a negative impact on themselves or their family in the future (62%). Research Co. President Mario Canseco noted that two-thirds of Canadian women (67%) foresee a negative personal or familial impact from AI, compared with 58% of Canadian men.
Trust Levels Vary by Institution
Most Canadians say they are finding it harder to trust content they encounter online, with 80% saying they have more difficulty trusting images due to the possibility of AI generation, and 77% saying the same about information that may not be verified reporting.
When it comes to who Canadians trust to develop and manage AI responsibly, doctors and nurses top the list at 59%, followed by universities at 56%. The federal government earned 45% trust, up seven points, while provincial governments earned 43%, up five points. Trust levels were lower for tech executives (38%), international governments (33%) and business executives and CEOs (31%).
Despite the rising unease, AI use remains widespread: 59% of Canadians say they have used AI tools such as ChatGPT, Claude, Google Gemini or Grok in the past month. Among users, the most common applications are writing and editing (40%), translation (33%) and image creation or modification (31%). Fewer respondents reported using AI for trip planning (25%), task planning (23%), meal planning (22%), exercise planning (21%), note organization (19%), home design (17%) or coding (9%).
The survey was conducted online from September 10 to September 12, 2026, among 1,008 adults in Canada, with results statistically weighted to match census figures for age, gender and region. The margin of error is plus or minus 3.1 percentage points, 19 times out of 20.
This article references reporting from:
