TRADESAFE survey: only 18% of US workers say their workplace has a policy on accountability for AI errors in safety documents
Key points
- Only 18% of surveyed workers say their workplace has an accountability policy for AI errors in safety content
- 51% of respondents who know how safety documents are made say generative AI is used to draft them; 49% of those say outputs are independently reviewed
- Leaders (74%) were far likelier than safety professionals (23%) to report independent review
A TRADESAFE survey of 514 US workers in manufacturing, construction, warehousing, logistics and skilled trades, conducted in September 2026, found only 18% say their workplace has a policy defining accountability for AI errors in safety content. Among 399 respondents, 51% said generative AI is used to draft safety documents, and 49% said outputs are independently reviewed. Leaders (74%) were far likelier than safety professionals (23%) to report review.
Original report: Human Resources Director ↗AI-assisted brief; may contain inaccuracies. Rely on the original reports and official documents. For information only; not legal, immigration or insurance advice.
