The Ethical Case for Using AI as a Workplace Safety Enhancement
AI adoption isn’t just a business imperative, it’s a moral imperative for improving worker safety.
- By Julia Penfield
- Mar 21, 2025
Much of the conversation surrounding artificial intelligence technology focuses on AI adoption as a business imperative. If companies don’t take advantage of AI, analysts and executives often warn, they’ll fall behind their competitors, who will leverage AI to gain efficiency and increase profits.
That’s one talking point in favor of AI adoption, but I’m here to offer a complementary view. To me, using AI to improve worker safety isn’t just a business imperative. It’s a moral imperative.
Indeed, for businesses concerned about improving safety outcomes for employees, AI doesn’t merely offer a way of speeding up safety management processes or boosting the productivity of safety teams. More importantly, it presents an opportunity to reduce risk and ensure that more workers can go home at the end of each day injury-free.
Allow me to explain by walking through the challenges businesses face in the realm of workplace safety and the unique role that AI can play in mitigating them.
Why we need AI to improve workplace safety
Viewed from one perspective, organizations have done a great job of improving workplace safety over the past several decades, starting well before the advent of AI-powered workplace safety technologies. The rate of worker injuries and illnesses declined by about 80 percent between the 1970s and the early 2020s, thanks largely to investment in better workplace safety controls and improved risk analysis practices.
Viewed from another perspective, however, workplace safety remains a deep and pressing challenge. That’s because if you look at the rate of serious injuries and fatalities (SIFs), you’ll notice that it has barely budged over the past several decades.
This means that, although workers today are less likely to suffer minor injuries on the job than they were in the past, their chances of experiencing serious injuries – the ones that cause grave harm, including death – remain virtually unchanged.
The stubborn frequency of SIFs is not due to a lack of interest or effort among employers in building a safer workplace. Instead, it’s a product primarily of challenges like budgetary constraints and a lack of workplace safety specialists, both of which have made it hard for organizations to maintain adequate staffing levels for safety professional teams. The increased administrative load placed on safety professionals as a result of growing regulatory and reporting requirements is a factor, too, because it has reduced the time that these professionals can devote to the work that matters most for safety outcomes – assessing and mitigating risks for workers.
Against this backdrop, AI offers tremendous promise as a means of finally moving the needle on serious workplace injuries in a positive direction. This is why adopting AI is an ethical and professional imperative for any business committed to keeping workers safe. To fail to take advantage of the opportunities AI presents in this area is to accept the status quo in which workers continue to experience serious injuries just as frequently today as they did a half-century ago.
How AI can enhance workplace safety
At a high level, the role that AI stands to play in improving workplace safety is that AI tools can help overstretched safety professional teams do more with less. In this way, AI mitigates the bandwidth limitations that safety professionals face as a result of staffing limitations and competing demands for their time.
To be more specific, AI can assist with a variety of common tasks and requirements for safety professional teams in ways that not just increase the efficiency of risk analysis and mitigation, but also translate to safer workplaces.
Risk identification assistance
Identifying risks and finding ways to remediate them are the most important parts of a safety professional’s job – but when safety teams are overstretched, they often can’t devote as much time and effort to these tasks as they ideally would.
For instance, safety specialists typically have little time to devote to reviewing every single incident report. As a result, they may overlook near-miss instances where no serious harm occurred, but where a future worker may not be so fortunate unless the business implements safety controls to mitigate the relevant risk.
AI can help by assessing potential serious injury or fatality (PSIF) incident reports and automatically identifying the risks that nearly caused an injury. It can also recommend safety controls that will prevent a similar incident from recurring.
AI-assisted Job Safety Assessments
In a similar vein, a best practice for minimizing the risk of injury before beginning a task is for workers and supervisors to perform a job safety assessment (JSA). Ideally, they’d do this by consulting with safety professionals, who can explain the risks and measures that employees should follow to mitigate them.
But given the overstretched nature of safety teams, professionals aren’t always available to offer JSA guidance. AI, however, can help by automatically generating insights about hazards and the safety controls available to protect against them. In this way, AI can function essentially as a virtual safety professional that is available 24/7 to any worker or manager seeking expert guidance on risk management. It can also help reduce the total rate of SIFs, leading to valuable improvements in workplace safety.
Worker risk assessment
AI can help as well in automating the process of detecting and assessing risks that workers face in the course of carrying out routine tasks.
For instance, by analyzing a video of a worker performing a repetitive task, AI could detect ergonomics related risks that might lead to eventual musculoskeletal/chronic injuries, as well as suggest mitigations. Capabilities like this would not only help to scale risk identification. They can also reduce challenges businesses face due to a shortage of certified ergonomists, which has limited employers’ ability to perform ergonomics assessments manually, while also reducing the rate of musculoskeletal injuries experienced by workers.
Post-incident root cause assessment
After an incident occurs, it’s imperative for the safety team to understand the root cause of what went wrong and how the risk can be mitigated in the future using Correction Actions. Overstretched safety teams, however, often struggle to perform this analysis manually and delegate it to trade workers and managers. And even when safety professionals do examine reports, it’s not always easy to determine what the root cause of the incident was. Individuals may operate based on certain assumptions that cloud their ability to assess risks objectively, for example, or they may overlook key data when reviewing an incident.
Well-trained AI models, however, are not subject to efficiency or visibility limitations. When AI tools can analyze all relevant safety data, they can home in on the key underlying root-cause risks that businesses must mitigate to improve safety. They can also identify which safety controls are most effective for minimizing those risks and recommend new safety Corrective Actions to implement – leading ultimately to fewer SIFs.
Organization-wide risk analysis
Many Large enterprises operate multiple sites, with different safety teams deployed at each one. The information that each team collects tends to exist in a silo, with the result that it can be challenging for the business to gain a holistic view of where its safety risks lie and which controls can best mitigate them. A team at one site might lack a fleet-wide point of view into which hazard types and energy sources are responsible for which percentages of PSIF and SIF incidents, for example.
AI can help here by systematically analyzing safety data from across the company, determining where the most significant risks lie and identifying which measures are most effective at mitigating them. This approach not only helps safety teams work more efficiently by sharing insights and ideas, but also improves outcomes for workers by ensuring that the best safety controls known to the company are deployed everywhere relevant.
Conclusion: Using AI to create value for workers, not just business
In conclusion, improving corporate balance sheets shouldn’t be the only reason why businesses leverage AI in the realm of workplace safety. What really makes AI critical is its ability to keep workers safer by enabling safety enhancements that overstretched professionals just can’t implement manually at scale.
This is the most important reason why I hope we’ll see more and more businesses taking advantage of AI-powered workplace safety tools in the near future. Saving money is great, but saving workers from serious harm – something that the past half-century of safety investments haven’t done all that well – is the utmost important value of using AI for safety, making it a moral and professional imperative for EHS software companies.