Risk Talk Spikes; Only 40 Safety Roles as Engineering Leads Hiring

An avalanche of AI risk talk is colliding with a clear hiring reality: only 40 safety roles in our data while engineering dominates. Alarms about extinction risk and agent attacks are louder, yet employers still prioritize engineering over safety, with OpenAI and Anthropic hiring strongly. Across the 193 employers we track, 4,464 AI roles were open as of 2026-09-12, down 114 on the week. Our model puts new AI jobs created month-to-date at roughly 66,125. A second look at postings echoes the split: employers list 4,685 AI vacancies across 178 firms, with just 40 in safety, supporting diffusion claims but not a slowdown. The through line is simple. Work is spreading into more teams and products, but dedicated safety headcount is not keeping pace. For engineers and product builders, that means more opportunity and likely more responsibility for risk mitigation inside core roles. For people aiming at safety, it means fewer openings and stiffer competition. For workers watching the boycott calls and fear narratives, the hiring boards tell a cooler story. The market is rewarding deployment and product work first, while safety hiring lags behind the rhetoric.