Industrial worker using AI chatbot on smartphone

Unsanctioned AI Use Rising Among Global EHS Teams

Research shows safety professionals are increasingly turning to unofficial tools as organizations struggle with data integration and governance.

New research reveals that artificial intelligence adoption in EHS sectors is outpacing organizational readiness, with 95% of senior leaders reporting that employees are using unapproved AI tools at work.

According to The State of EHS+ Technology report, this rise in "shadow AI" occurs as organizations struggle with fragmented data systems. While 97% of leaders acknowledge AI use within their operations, only 5% have successfully embedded the technology across their workflows.

The study, which surveyed 2,000 senior leaders across North America and Europe, highlights a significant infrastructure gap. Approximately 85% of organizations still rely on manual or disconnected tools. Furthermore, 88% of respondents noted that a lack of integration between EHS software and other enterprise systems remains a primary barrier to scaling AI effectively.

Despite these hurdles, AI is significantly influencing procurement strategies. The data shows that 94% of leaders consider AI capabilities when purchasing new technology, and 62% are moving toward integrated platforms to consolidate their data.

The report suggests a shift in how safety professionals view the value of AI. Rather than using it solely for administrative tasks or backward-looking reports, 82% of organizations see the potential for AI to assist in real-time decision-making. This includes managing shift changes, monitoring PPE availability and identifying worker fatigue as it happens.

Trust remains a critical factor in adoption. Only 15% of respondents said they would trust fully autonomous AI to make safety decisions. Instead, 85% of leaders prefer a "human-in-the-loop" approach, where AI provides recommendations that are ultimately reviewed and approved by a person.

The findings indicate that the most "mature" organizations are not necessarily those with the most AI tools, but those that have prioritized centralized data and integrated platforms as a prerequisite for automation.

About the Author

Jesse Jacobs is assistant editor of OHSOnline.com.

Featured

Artificial Intelligence