AI’s quiet safety gatekeepers are stepping into the spotlight

The intensifying AI safety debate is bringing a small group of third-party evaluators into the center of a multitrillion-dollar industry.