How “trace hiring” can overcome the deluge of AI job applications

A single job posting today can pull in 150 to 200 applications within its first 24 hours, most of them shaped, in part, by AI writing tools. Recruiters are drowning, trust between candidates and employers is fraying, and keyword-matching software cannot tell a rehearsed answer from a genuine one. In my new piece for LSE Business Review, I propose a way out: Trace Hiring.

Trace hiring borrows its logic from stigmergy, the same coordination principle that lets termite colonies rebuild an entire nest with no blueprint and no foreman, first documented by entomologist Pierre-Paul Grassé. Instead of scoring resumes, trace hiring asks candidates to act, and then reads the traces that action leaves behind: how they respond under pressure, how they behave in an unscripted group exercise and how honestly they can reconstruct what just happened. The article lays out a three-stage process: mass filtering through a single hard question, unmediated group observation and a final reflective interview. This process is designed to let real judgment and agency surface on their own, without coaching, scripts or AI polish getting in the way.