Workplace safety leaders are adopting AI, but warn guardrails are needed

The AI Evolution in Workplace Safety: Moving Beyond the Pilot Phase

For years, artificial intelligence (AI) in environment, health and safety (EHS) was treated as a futuristic experiment—a series of pilot projects and “what if” scenarios. But, the landscape is shifting. AI is rapidly moving from the testing lab into everyday operational use.

According to The Safety Shift: EHS Readiness in 2026, a report based on survey responses from 1,053 professionals in EHS, operations, and risk management across U.S. Industries, the technology is now being applied to core safety functions. This includes incident prediction, hazard identification, regulatory compliance, analytics, and reporting.

The adoption curve is steep. While 20% of organizations report extensive AI application within their EHS programs, another 62% report moderate or limited use. This suggests that AI is no longer a niche tool for tech giants but a mainstream expectation for safety leaders.

Bridging the “Digital Divide” to Unlock AI Potential

Despite the enthusiasm for AI, there is a significant hurdle: the underlying data infrastructure. AI is only as effective as the data that feeds it, and many organizations are still struggling with fragmented systems.

The data reveals a stark reality regarding digital transformation. Only 11% of organizations have fully digitalized EHS systems. A vast majority—71%—operate in hybrid environments that blend digital tools with manual, paper-based workflows. Another 18% still rely primarily on manual processes.

This reliance on hybrid or paper-based systems creates a bottleneck. When safety data is trapped in spreadsheets or physical folders, AI cannot analyze it in real-time to predict hazards or identify trends. The “digital gap” is most evident in day-to-day operations, such as permit-to-work systems, occupational health monitoring, and behavior-based safety observations, where digitalization lags furthest behind.

The Human-in-the-Loop: Balancing Automation with Judgment

As AI takes over more analytical heavy lifting, a new risk is emerging: overreliance. The transition to AI-driven safety is not without anxiety, as 90% of safety professionals report at least one concern regarding the technology.

The primary fear is the loss of human intuition. Roughly 65% of professionals cite overreliance on AI as a key risk. The consensus among industry leaders is that while AI can identify a pattern, it cannot replace the nuanced judgment of an experienced safety professional on the shop floor.

Maintaining a “human-in-the-loop” approach is essential. This means using AI to flag risks and provide data-driven insights, while leaving the final decision-making and accountability to humans. This balance ensures that safety remains a proactive human responsibility rather than a reactive algorithmic output.

Expanding the Perimeter: Mental Health and Psychosocial Risk

The definition of “workplace safety” is undergoing a fundamental expansion. Safety is no longer just about hard hats and fall protection; it is increasingly about the mind.

An overwhelming 87% of professionals agree that mental health belongs within the scope of EHS. This shift recognizes that psychosocial risks—such as chronic stress and burnout—can be just as detrimental to worker safety as physical hazards.

However, there is a gap between recognition and immediate priority. While mental health is accepted as part of the mandate, near-term operational priorities still lean toward fatigue management, infectious disease preparedness, and the challenges of managing an aging workforce.

Jay Vietas, Senior Director of Research at the National Safety Council, emphasizes that EHS leaders are in a “period of significant transition,” working to balance these new workforce risks with longstanding operational responsibilities.

Frequently Asked Questions

Is AI replacing safety professionals?

No. The trend is toward “augmented intelligence,” where AI handles data analysis and prediction, but humans provide the critical judgment and accountability necessary for safety decisions.

What is the biggest barrier to AI adoption in EHS?

The lack of a fully digital foundation. With 71% of organizations still using hybrid digital-manual workflows, inconsistent data prevents AI from reaching its full effectiveness.

Does mental health fall under EHS responsibilities?

Yes, 87% of surveyed professionals agree that mental health is part of the EHS scope, though it often competes with immediate priorities like fatigue management.