Ethical AI-Use Guidance for Workplace Health and Safety Management  

Summary: An expert on applied ethics describes how employers can use a framework to evaluate AI-related exposure risks and make ethical and legally defensible workplace health and safety management decisions.   

AI is penetrating all aspects of daily life at a breath-taking pace. In the workplace, AI applications that streamline processes and expedite production are a primary focus. But AI implementation is not just a technological exercise. It requires thoughtful leadership to make complex ethical decisions.  

There is a framework employers can use to reach legally defensible ethical decisions when using AI to enhance occupational health, safety, and risk management programs. During a recent WorkCare webinar, Mark Katchen, founder of Applied Ethics Advisors, describes how an “ethics analyzer” model helps reduce the chance of being caught off guard by AI-related ethical dilemmas.  

About the Process 

Katchen applies 40 years of experience as a certified industrial hygienist and safety professional, educator, author, and consultant to the resolution of murky ethical issues. He recommends the adoption of a robust review and documentation process, including audits of routinely used AI applications like wearable tracking devices, ChatGPT, Claude, and Co-Pilot.  

Ethical risk-analysis involves the creation of a concise (one-page) decision-making template that can be used to: 

• Audit AI tools and their specific features 

• Record input from stakeholders 

• Explain intended use 

• Assess limitations 

• Identify legal and ethical concerns 

• Explain the rationale for each ethical conclusion 

“You’re already making AI-related ethical decisions that will be judged in hindsight by regulators, courts, workers, and your own leadership team. OHS professionals often own these decisions without AI utilization rules to follow,” Katchen said. “An AI mistake used to be a software bug. In our field, it’s now an incident investigation, a citation, or a deposition. What protects you is evidence that you saw the trade-offs and documented your thought processes.”  

AI, Ethics, and Hazards in the Workplace 

Thoughtful ethics guardrails have been developed. However, from a practical standpoint, AI is advancing so quickly that it can feel like the horse galloped away from the barn long before end users could grasp its magnitude. 

A global framework adopted by UNESCO member states defines AI-related ethical practices as “respect for human dignity” that are grounded in principles such as transparency, fairness, environmental sustainability, and human oversight. Other organizations have adopted similar statements. In its AI Risk Management Framework, the National Institute for Standards and Technology identifies key components of trustworthy and ethical AI as:  

• Valid and reliable 

• Safe 

• Secure and resilient 

• Accountable and transparent 

• Explainable and interpretable 

• Privacy-enhanced 

• Fair, with harmful bias managed 

Specific to U.S. workplaces, the National Institute for Occupational Safety and Health (NIOSH) defines AI as a potential hazard that can be identified by adapting established identification and exposure assessment methodologies to its “novel characteristics.” For example, NIOSH cites an article published in the American Industrial Hygiene Association’s journal, The Synergist, which suggests linking algorithmic hygiene systems – including ethical considerations such as worker-management trust and the presence of psychosocial hazards – to employee health and safety outcomes to help reduce exposure risks.  

A related NIOSH blog post states that ethical guidelines in the workplace should be aimed at ensuring AI systems: 

• Respect for autonomy 

• Safeguard fairness in operations 

• Promote the benefits of AI 

• Prevent harm 

Katchen said NIOSH’s framing of AI as a risk helps inform ethical decisions being made by OHS professionals who already know how to identify, assess, and control hazards.  

Looking Through Ethical Lenses 

According to Katchen, it’s essential to understand the difference between compliance (allowed or not allowed) and ethics (wrong, right, defensible). He said most ethical quagmires are not caused by a crisis but by “ethical drift” that occurs when an exception becomes a norm, competing obligations arise, and authority pressures delay action.  

Six ethical lenses are typically used to view a given situation: utilitarian, rights, justice, common good, virtue, and care. In addition to ethical considerations, OHS professionals are also expected to abide by certain codes of conduct, support risk-and-consequence analyses, and navigate organizational and authority dynamics.  

When under pressure, Katchen said there is a tendency to default to the utilitarian lens, which focuses on efficiency, even though all lenses must be applied to produce a defensible position. Each lens has a specific audience, for example, rights fall into the regulator domain; justice is relevant to a plaintiff’s attorney. The decision-making framework involves use of each lens and an algorithm to document and operationalize findings. 

To use ethical lenses, follow these steps: 

1. View the ethical issue in question through each lens. 

2. Use a weighted 1-10 scale to take a position on each point of view. 

3. Apply consistent weighting. 

4. Consider patterns that develop – higher versus lower weight. 

5. Do not average the scores. 

6. Document all the scores to create a record of what you believed and why. 

7. Test decision stability. How would you defend your decision if challenged? 

8. Take remedial action when an ethical breach is detected to avoid violations. 

 
“This process does not supplant your professional judgment. It provides guidance on how to proceed on a rational basis,” Katchen said. “It produces a record you can hand to a regulator or a leadership team to give them a clear picture of your rationale as you went through the process.” 

What Can You Do Next? 

One way to gain employee buy-in for AI-assisted applications is to explain the rationale behind their adoption and how the workforce will use these tools. “Transparency is key with most ethical issues. When you fail to be transparent and establish trust, that’s where failures occur,” Katchen said. 

The overarching objective is to apply a rigorous and consistent process to decision-making. Leaders are likely to be judged not only based on the decision itself but by whether they can explain how it was reached. Employers are also encouraged to familiarize themselves with proposed and existing AI rules that apply in their operational jurisdictions. Legal and regulatory guardrails for use of AI in workplaces vary by state. In some cases, they do not exist. (Refer to the Carlton Fields AI Employment Law Map for state-by-state status.)  

While waiting for regulatory agencies and courts to provide direction in this rapidly evolving environment, it’s a best practice to conform with ethical standards and continue to make workplace healthy and safety a high priority. Contact WorkCare to learn about the ways we assist employers with the delivery of preventive interventions and injury management solutions with the well-being of employees and best business practices at the forefront.