High-stakes AI: Why unreliable AI decisions are too risky for high-hazard industries
As global regulations shift from voluntary ethics to enforceable safety laws, the era of the ‘black box’ is over. Pulkit Parikh, PhD., machine learning scientist at VelocityEHS, discusses with us why transparency, auditability, and human-in-the-loop design are the new foundations of worker protection.
By Pulkit Parikh, VelocityEHS speaking to Safety Management on 16 March 2026