A new artificial intelligence system developed by researchers can analyze short speech samples to predict the progression from mild cognitive impairment to Alzheimer’s disease with high accuracy. The model, trained on audio recordings and corresponding transcripts from hundreds of participants in a long-term aging study, identifies subtle linguistic changes imperceptible to human listeners. It focuses on aspects like word choice, syntactic complexity, and acoustic features such as speech pace and pitch. In validation tests, the AI significantly outperformed standard clinical assessments at forecasting which individuals would develop Alzheimer’s within six years. The researchers emphasize the tool is intended to assist, not replace, clinicians, potentially enabling earlier interventions. The full study details are available in the published research article.