Wearable AccuracyAlgorithm BiasWomen Health
Your Smart Ring Was Trained on Men: The Gender Data Gap in Wearable Algorithms
Most wearable algorithms are trained on datasets that are 70 to 80 percent male. Heart rate thresholds, sleep stage classifiers, and stress baselines are systematically less accurate for women. Here is why, and what Pulsyn is doing about it.
James Hoffmann Jun 18, 2026 · 1 min read