Eli Lilly and Company has licensed the Alzheimer’s Drug Discovery Foundation (ADDF)’s SpeechDx dataset to study whether recurring voice samples paired with health records can support Alzheimer’s research.
ADDF announced the license Wednesday, Oct. 28, and said Lilly is the third organization to use SpeechDx. Siemens Healthineers and Callyope were named in earlier foundation announcements. The arrangement gives Lilly access to a resource built through ADDF’s Diagnostics Accelerator, but the announcement did not describe the analyses Lilly plans.
ADDF says the three-year program collects quarterly speech samples from more than 2,000 participants and pairs them with clinical and biomarker records. These include neuropsychological tests, MRI scans and blood-based amyloid status. Many participants also have PET imaging and cerebrospinal fluid data.
Public sources give differing participant totals. Its current website lists 2,650 people, while a June 2026 Perspective in Alzheimer’s & Dementia describes 2,006. The sources do not explain the difference.
The authors of the June Perspective said speech research has been constrained by a lack of large, multilingual collections that repeatedly capture voice and pair it with detailed health information. The article discusses the rationale for SpeechDx; it is not a report of results from Lilly’s work.
ADDF says recordings are encrypted, personal information spoken during them is manually removed, and the resulting voice data are pseudonymized. The foundation also notes that a person’s voice may still be identifiable. Approved researchers use the data in a controlled workspace where files cannot be downloaded or exported.
ADDF did not disclose the license’s financial terms, a schedule for Lilly’s research or a date for findings. Its announcement reported no results showing whether speech measures can diagnose Alzheimer’s disease or predict disease risk.
ADDF’s current project page anticipates completion in the fourth quarter of 2027. A 2025 foundation announcement had said data collection was on track to finish in 2028; neither announcement specified a milestone for Lilly’s research.
