A cross-sectional study of 903 devices published in JAMA Network Open in April 2025 found that less than one in three FDA-authorized AI medical devices have been validated on sex-specific patient populations, and the FDA's current rules let developers reach the US market without that validation. A new analysis published September 17, 2026 by The Conversation traces the problem from post-war regulatory decisions that barred women from drug trials to the AI systems now learning from that clinical record. The FDA authorized 168 machine-learning-enabled Class II medical devices in 2024 alone, a record year, yet a 2024 ML-device regulatory analysis found only 15.5% of those devices provided any demographic performance data and only 29.2% reported both sensitivity and specificity. Across 903 FDA-authorized AI devices through August 2024, less than one-third provided sex-specific data in clinical studies and fewer than one in four addressed age-related subgroups. In almost one-quarter of submissions, sponsors told the FDA that no clinical performance study had been conducted at all. The fast-track 510(k) pathway does not require prospective clinical studies. A separate JAMA Alzheimer AI devices study published in July 2025 confirmed limited demographic data transparency for devices authorized for Alzheimer's disease and related dementias.