Artificial Intelligence (AI) is increasingly being used in Canadian workplaces and industries to augment or enhance work, or to perform certain job tasks, previously performed by workers.
But to what extent is AI being used to help keep workers safe on the job?
According to an Institute for Work & Health (IWH) study that surveyed 810 occupational health and safety (OHS) professionals from British Columbia (BC) and Ontario, about a quarter said that their organizations used AI for health and safety purposes. About half said their organizations did not use AI for OHS, and the remainder of participants did not know.
Large organizations and organizations with a high number of hazards were the most likely to report using AI for OHS. Participants with more positive perceptions of the utility of AI for OHS also were more likely to report that their organizations used the technology for this purpose.
The study, led by IWH Associate Scientific Director and Scientist, Dr. Arif Jetha and published in the American Journal of Industrial Medicine (doi:10.1002/ajim.70018), asked OHS professionals about the use of AI in their organizations, their personal perceptions of AI, and characteristics of their organizations.
“With the rapid advances in AI, we wanted to know more about whether and how organizations were using AI for health and safety purposes, and which organizations were more likely to use it,” says Jetha, who also presented preliminary findings from this study at an IWH Speaker Series presentation earlier in 2025.
Respondents from organizations that used AI for OHS said they used it for a range of applications. Twenty-seven per cent used AI to identify and eliminate hazards (e.g., with AI-powered fire suppression systems), 22 per cent used it to monitor safety behaviours in combination with cameras and sensors, 19 per cent used it to promote health‐enhancing behaviors (e.g., with mental health monitoring tools), 15 per cent used it to help prevent psychological injuries (e.g., with AI therapy apps), 10 per cent used it to help prevent physical injuries, and 7 per cent used it for disability management.
“Given how much we hear about AI and its various uses, AI use for OHS was not as common as I anticipated among this sample,” says Jetha. “But as AI becomes more affordable and widespread, new OHS tools will likely be developed. We might see an increase in these numbers. It also will be important to monitor the efficacy of AI tools for OHS, as we don’t yet know how effective these AI tools are in preventing workplace injuries or illness.”
Which organizations are most likely to use AI for OHS?
Compared to organizations not using AI for OHS, those that did use it were 2.5 times more likely to be a large organization (with 500 or more employees) and 1.7 times more likely to be a medium-sized organization (between 100 and 500 employees). Organizations that used AI for OHS were also 1.7 times more likely to be based in Ontario than BC. The researchers found no differences in use between industries in this sample.
Organizations using AI for OHS were also more than three times as likely to be highly hazardous workplaces (workplaces where workers were frequently exposed to hazards) and twice as likely to be moderately hazardous compared to organizations not using AI for OHS.
“Our finding that larger organizations and those with potentially more hazards report greater AI use for OHS may highlight need and resource differences in firm-level AI uptake. The most safety-sensitive and well-resourced organizations may be leading the way in turning AI to innovate health and safety practices,” says Jetha. “It could also be that large organizations have an existing technological infrastructure that may make AI adoption more straightforward.”
OHS professionals’ perceptions of AI
Participants from organizations using AI for OHS were more than three time as likely as non-users to have positive perceptions about the impact of AI on creating safer workplaces, addressing OHS costs and reducing psychological and physical injuries.
“AI users might be more knowledgeable about the pros and cons of the and have had more positive experiences,” says Jetha. “What this tells us is that as the AI landscape expands, it will be important for OHS professionals to have more information about AI, gain opportunities to experience its potential benefits, and think about drawbacks so they can contribute to the design and adoption of the technology in their organizations.”
How was the study done?
A total of 810 participants in British Columbia and Ontario were surveyed in the summer of 2024. All participants had OHS responsibilities. Most were supervisors, senior managers or executives; others included OHS practitioners, human resources professionals or union representatives.
Surveys asked about whether the organizations they worked for used AI for OHS, along with characteristics about their organization, including: the number and gender split of employees; whether employees primarily worked remote, in-person, or in hybrid arrangements; and whether a union represented any staff.
Study participants were also asked about whether they felt the OHS policies of their organization were adequate, and how frequently their workers might be exposed to hazards or unsafe working conditions. Finally, they were asked about whether they had concerns about AI, or thoughts on its potential impact for health and safety.
Do these AI tools make an impact on worker health and safety?
The effectiveness of AI programs to improve health and safety is not yet clear—as Jetha’s systematic review of existing research found.
“We’re still at the early stages of understanding the role AI can play for OHS and building tangible recommendations for OHS professionals,” says Jetha. “We need to look into the utility and effectiveness of AI for OHS problems across different types of organizations.”
To better address many of these issues, Jetha was recently awarded a seven-year research partnership grant called the Partnership on AI and the Quality of work (PAIQ), which will dig further into the role AI can play in OHS. Stay tuned as findings from that project develop.