As AI enters workplace safety decisions, a certification leader explains why judgement still matters most
Competence in occupational health and safety used to mean years on the job and a few hard-won lessons. Today, it also means knowing when to question an answer generated by artificial intelligence. That is the central argument from Larry Masotti, CRSP, chair of the governing board of the Board of Canadian Registered Safety Professionals (BCRSP), who says the profession’s third-party certification system matters more, not less, as AI tools spread through safety decision-making.
“We all benefit from theoretical and academic knowledge and practical and technical experience, but we always need to ensure that ongoing development or experiential learning is happening,” Masotti said. “The BCRSP has published a competency framework which is available not only to certificants but anyone that goes online to look at it, and it basically is an objective assessment of technical, behavioral, and leadership competencies.”
The three lenses behind every safety hire
Masotti said employers typically judge occupational health and safety through three lenses, and a gap in any one exposes a hidden assumption about who counts as competent. “One is the legislative lens: are we meeting the intent of the law of the land, whether that’s provincial, territorial or federal,” he said. “The second piece might be the economic impact of a health and safety incident, and that could be presenteeism, where people are not engaged in the workplace, they are simply putting in time. And a third would be perhaps the moral aspect: are we doing the right thing? Are our employees engaged?”
Taken together, those three lenses connect day-to-day safety competence directly to enterprise risk, since a gap in any one can expose an organization well before a serious incident occurs.
Why AI raises the stakes for sound judgement
Masotti was direct about both the appeal of AI tools in safety work and the risk that comes with it. “The attraction of AI is the fact that it is novel, and it does some amazing things,” he said. “So it is enticing, it is fascinating, and it’s concerning, because when solutions are not achieved through a reflective exercise, people coming into the field do not have the necessary background to determine if the information they’re getting fits the problem they’re presented with. It’s developing so quickly that I don’t know that we know exactly all the challenges we’re going to face in this regard.”
Asked how an employer can tell whether a safety leader has the judgement to know when to trust AI output and when to challenge it, Masotti pointed to a specific discipline. “I would suggest one of the fundamental pieces in considering occupational health and safety is to maintain a healthy skepticism, meaning always questioning the source, always questioning the manner in which the problem is approached, to determine have we considered all things,” he said.
Reverse onus puts the burden on employers
That skepticism is not optional, Masotti said, because of a legal principle that shapes every health and safety decision in Canada. “All health and safety decisions are predicated on the concept of reverse onus, meaning that an employer or worker or supervisor is guilty until proven innocent,” he said. “That is the reverse of the normal procedural law, where I’m innocent until proven guilty. How do you prove your innocence? Well, of course, you have been duly diligent and you have done everything reasonable under the circumstances.”
Under that standard, a solution pulled from an AI platform carries the same burden of proof as any other safety decision, and that burden cannot be delegated away, Masotti said. “The onus is on the employer, which he or she cannot delegate to a health and safety person,” he said. “The law indicates that we are all responsible for health and safety in an organization, so there has to be some rigorous testing to ensure that the solution is sound.”
Independent verification, he said, can mean checking the original source, consulting a subject-matter expert, reviewing historical references, and running the decision through a documented assessment matrix that records the rigor behind the final call.
Certification as third-party proof
That is where certification fits in. Credentials such as the Canadian Registered Safety Technician (CRST) and Canadian Registered Safety Professional (CRSP) designations give employers an objective, third-party check on a candidate’s technical, behavioral and leadership competencies, validated against ISO/IEC 17024, the International Organization for Standardization’s standard for certifying individuals.
“You have an objective third-party validation of competence,” Masotti said. “The competencies and the domains which are tested parallel the continuing evolution of health and safety in our workplaces. Years ago we talked about slips and falls and cuts and bruises, but today we’re talking about a wider spectrum of risks and hazards, AI being one of them, mental health being another.”
That widening scope reflects BCRSP’s own expanding mandate as the field grows to cover psychological safety and AI, a shift that also runs through the wider debate over the privacy risks and regulatory gaps surrounding new safety technology. Masotti said the competency framework is a public, downloadable document on BCRSP’s website, meant as a reference point for employers working out what sound judgement looks like in an AI-assisted workplace.
