Eli Lilly Licenses Largest Speech Dataset for Alzheimer's Research

What the Lilly licence covers

The Alzheimer's Drug Discovery Foundation (ADDF) and its Diagnostics Accelerator (DxA) announced the collaboration on Oct. 28, 2026. Lilly will use SpeechDx to investigate the usefulness of speech in Alzheimer's research.

The ADDF describes the dataset as the world's largest harmonized longitudinal collection of speech and biomarker data across the Alzheimer's continuum, from cognitively unimpaired individuals to people with Alzheimer's dementia. Lilly becomes the third organization to license it, after Siemens Healthineers and Callyope.

Mark Mintun, group vice president for Neuroscience Research and Development at Lilly, said the dataset pairs speech-based data with established biomarkers. He said this supports the company's efforts "to make diagnosis more accurate, accessible, and less invasive across the disease continuum." He added that earlier detection is critical because mild symptoms are often mistaken for normal aging, which delays opportunities for intervention.

How SpeechDx was built

The DxA launched SpeechDx in 2023. It is the ADDF's largest investment in AI to date, according to the foundation. The project targets a gap the ADDF describes as critical: the lack of datasets that track how a person's speech changes over time alongside biomarker-confirmed diagnoses and clinical outcomes.

The study follows more than 2,000 participants over three years. The ADDF calls the result a gold-standard resource that AI models can train and validate against.

The foundation argues that speech is one of the earliest indicators of cognitive decline. Because it can be captured inexpensively through the microphones already built into smartphones and computers, it offers what the ADDF calls a highly scalable way to monitor brain health.

Why data quality matters for digital markers

ADDF Chief Executive Officer Isobel Coleman said AI-driven tools have the potential to transform how Alzheimer's is detected and monitored, but "they are only as good as the data behind them." She said SpeechDx is designed to give researchers and industry partners what they need to develop and validate the next generation of digital markers.

Niranjan Bose, Managing Director, Health and Life Sciences at Gates Ventures, said the field needs large, well-characterized datasets to turn the speech signal into a reliable digital measure. He said a group of high-caliber researchers has already been convened alongside partners such as Siemens Healthineers and Callyope, and that adding Lilly brings the effort closer to establishing speech as a credible, validated tool for Alzheimer's research and drug development.

The ADDF says the licences underscore the potential value of dynamic, digital measures used alongside blood tests to give additional information on a patient's cognition and future risk.

The organizations behind the effort

The ADDF was founded in 1998 by Leonard A. and Ronald S. Lauder and says it is the only public charity solely focused on funding drug development for Alzheimer's. It reports having awarded nearly $400 million to fund 792 programs in 21 countries.

The Diagnostics Accelerator, launched in July 2018, is a $150 million global research initiative founded by Leonard A. Lauder and Bill Gates. Its aim is to develop novel biomarkers for Alzheimer's disease and related dementias, including peripheral and digital biomarkers.

What it means for research

The licence gives a large pharmaceutical company access to a dataset that links speech changes to confirmed diagnoses over time. The ADDF presents the deal as growing industry recognition of AI-driven speech analysis for predicting Alzheimer's risk. The source does not say what results Lilly expects or when, so the work remains at the research stage. For the field, the three licensees show that developers of diagnostics and therapeutics are now testing speech as a possible complement to blood tests.