Thoughts on “AI in OEHS”

Alan Leibowitz’s article “AI in OEHS: Ethics of the Future” in the September Synergist identifies ethical concerns with the use of artificial intelligence by occupational and environmental health and safety professionals. These concerns include data privacy, the possibility of bias in AI outputs, and the need for accountability and transparency as well as human supervision when AI systems are incorporated into OEHS practice.

As explained in the article, AI tools powered by large language models (LLMs) can respond to natural-language questions in many subjects, but these responses cannot be accepted at face value. These systems do not inherently possess “knowledge” or “facts” in the way humans understand them. An LLM-based tool is optimized to provide the most likely response—not the most correct one—based on its training data.

CASE STUDY A fictional case study presents the story of Tobor, a young industrial hygienist at a large environmental consulting firm. Assigned the tedious task of collecting historic corporate data for a new client, Tobor turns to AI to help prepare his report. He prompts the system to summarize the data with particular attention to applicable regulations and exposure hazards that require action. After a brief review, he submits the report to management, who base their recommendations to the client on it. But later, discussions with the client reveal that some of the standards identified in the report do not exist or were inaccurately applied to the issue being addressed.