It seems a safe bet that within a few short years, some degree of artificial intelligence (AI) will be found in just about every occupation, whether to improve efficiencies and enable better decision-making or better protect health and safety. Many occupational health and safety (OHS) applications for AI are already available, and use is growing. “As we’ve seen AI technologies mature, so have the benefits,” explains Dave Chalmers, product vice-president at SafetyCulture, a U.S. maker of workplace AI platforms. “These technologies boast not just speed and accuracy but also the ability to detect and predict, so organizations can take a proactive approach to risk management.”
The power of AI lies in its ability to monitor and analyze vast amounts of data. In workplace settings, this means it can monitor and analyze the entire health and safety picture – environmental, equipment and human behaviour data, from ambient gas composition and machine motor temperature to physical obstacles and signs of human fatigue. Then, as needed in real-time, the AI can alert workers to risks and even direct emergency response. When connected to cameras and sensors, AI can also analyze worker movements to detect things like improper equipment handling to prevent long-term strain injuries.
At the same time, AI oversight supports better compliance, notes Chalmers, reducing the risk of fines, legal battles and compensation, and damage to an organization’s reputation. AI can also enhance safety training through using virtual reality and other types of interactive learning, all in an environment free of the risks associated with actually handling hazardous materials or working at heights, to name just two examples. Using drones and computervision technology, advanced AI systems can eliminate the need for humans to be physically present in high-risk inspection scenarios. The applications really are endless.
Data is key
Of course, the more data AI has, the more useful it is. In the big picture of workplace health and safety, AI can identify patterns, point out if one shift or location is having more safety incidents than another (and provide some insight into why), highlight areas for improvement and much more. All of this information enables leaders to make the best decisions about updating safety protocols, investing in new tech and more. (At the same time, of course, data needs to be protected and used ethically.)
Because AI is only as good as the data it’s provided, Chalmers says that the best results are seen when these systems have complete visibility across an organization’s entire operations. “Businesses are often using multiple systems for different aspects of their safety management, creating complexity and a data disconnect,” he explains. “When safety management systems are integrated and speak to each other, that’s when organizations will unlock the most valuable insights and opportunities.”
In his view, leaders also need to execute AI rollout in tandem with people management, including education and training. “One approach we’ve seen work successfully, internally and with our customers, is identifying internal champions within teams that can support and advocate for the project from the outset,” Chalmers explains. “These champions … spot opportunities for technology to help workplaces run safer, smoother and better, connecting frontline workers and leaders, and rallying teams to drive change. They can work hand-in-hand with AI models to speed this up, but personal experience and expertise will always be critical to this process. Champions can make or break the success of a project, especially if there’s internal resistance to change, or a perceived ‘top-down’ rollout mandate.”
Factors affecting adoption
SukhDev Mishra at India’s National Institute of Occupational Health can provide more insight into what influences the integration of AI solutions in workplace health and safety. In his view, these are primarily the level of risk to which workers are exposed, the availability of reliable AI-driven solutions, and cost. He recently published an overview of AI implementation in workplace settings with his colleague, Immad Shah, in the Journal of Occupational Health.
These three factors are connected to specific industries. “Certain sectors such as agriculture, logistics and manufacturing are leading the way due to the relative ease of implementation and the availability of AI solutions tailored to their specific risks,” Mishra says. “While many companies express interest in adopting AI for hazard detection, predictive maintenance and behaviour monitoring, they are still evaluating their options to ensure minimal disruption to existing operations.”
Mishra also notes that, in general, the pace of AI development in occupational health and safety is relatively slower compared to high-priority domains like healthcare, finance and autonomous systems. AI solutions fundamentally rely on well-structured, high-quality datasets, he says, “Yet OHS data is often fragmented, inconsistent or lacks the granularity needed to develop robust predictive models. Unlike healthcare and finance, where standardized data collection has been in place for years, workplace safety data is still evolving, making AI adoption more challenging.”
Another major limitation to AI adoption in OHS is stiff competition for AI talent and computational resources, with these tending to go to industries that directly drive business growth like e-commerce and the development of new technologies for widespread sales to the public. “Workplace safety, while critical, often remains a compliance-driven function rather than a core investment area, which slows down funding and innovation,” notes Mishra. However, he also says AI in OHS is expected to gain momentum “as industries recognize its long-term benefits in cost reduction and efficiency improvements.”
A look at AI-enabled equipment
Among the AI-enabled devices now available, Mishra and Shah point to “smart helmets” equipped with a sensor array – for example GPS, RFID, ultra-wideband sensors, an around-view monitor and even air quality sensors. This means that instead of sensors being located in the worker environment, the worker’s wearable device can collectively monitor worker location, activities and even the surrounding environment for improved individual health and safety.
Battery-powered exoskeletons are another example of AI-enabled OHS devices. These wearable robotic suits enhance the function of muscles and joints, preventing injury while also boosting productivity. Germany-based German Bionics is a global leader in exoskeleton development. “Over the past two to three years, and especially into late 2024 and early 2025, we’ve observed a significant increase in interest … across multiple sectors,” reports Thomas Leliveld, director of operations for North America. “The most active industries remain logistics and manufacturing, as these environments are characterized by physically demanding,
repetitive tasks.”
He explains that the company’s exoskeleton provides appropriate external energy through actively learning and adapting to individual users’ movements. “Unlike passive or hybrid exoskeletons, which require users to expend energy to receive support, our system delays the onset of fatigue,” he says. “As a result, workers experience greater stamina and endurance, increasing their potential output while protecting their health and well-being. Handling tasks require less effort, making movements easier and often faster. Users benefit from up to 36 kg (80 lb) of support per lifting movement, active walking assistance and counterforce capabilities for protecting the back when lowering loads or working in bent-over postures.”
ROI calculations for these types of devices encompass injury rates, productivity gains and many other factors, but because jobs are made easier with these devices, easier recruitment and retention should also be considered.
Looking forward
Leliveld is very positive about the future of AI-enabled devices like exoskeletons in health and safety, predicting that they will play an increasingly critical role in individualizing support. “As the technology continues to advance, motion control, power assistance and overall comfort will become even more precisely optimized to meet individual needs,” he says.
Looking at AI in OHS in general, Chalmers says that the effectiveness of AI in workplaces will be realized in how much it enhances decision-making and other human capabilities. And while in his view, “there remains a high level of skepticism about AI amongst [health and safety] professionals that needs to be addressed,” he also believes “the role of AI in safety management is only going to become more prominent, and organizations that fail to embrace it will ultimately fall behind.”
Mishra predicts that over the next five years, AI adoption in OHS is likely to progress at different rates in different areas, depending on industry, regulatory frameworks and tech readiness. “Predictive analytics, hazard detection and real-time monitoring are expected to see increased use, particularly in high-risk industries like construction, mining and manufacturing, where AI solutions can offer measurable safety improvements,” he says. “However, adoption may be slower in sectors with lower perceived risk or constrained budgets, though advancements in cost-effective AI solutions and regulatory incentives could influence this trajectory.”
Over the next 10 years, he believes AI-driven automation, wearable technology and digital twinning could become more widely integrated into safety protocols, leveraging coming advancements in machine learning, the Internet of Things and robotics. “Industries that prioritize workforce safety as a key operational metric, such as logistics and hazardous environments, are likely to lead in adoption,” says Mishra. “Meanwhile, AI’s role in ergonomic assessments, mental health monitoring and adaptive safety systems could become more pronounced.”
Overall, he believes the rapid evolution of AI presents both opportunities and challenges in OHS. “The key to widespread adoption,” he says, “will be the development of scalable, reliable, and cost-effective solutions that can seamlessly integrate into existing safety protocols.”
Treena Hein is a freelancer writer based in Ontario.
