Muse Maker Used Users’ EEG Data to Train AI Without Explicit Notice

Summary
A year-long investigation found Interaxon used about 50,000 hours of Muse users’ EEG recordings to train AI without explicitly informing them. The company updated its privacy policy; research suggests EEG can carry re-identification risks.
Key points
- A year-long investigation reported that Interaxon used about 50,000 hours of users’ EEG recordings to train a brain foundation model without explicitly informing them.
- Interaxon removed names and email addresses from the training data and updated its privacy policy after Science questioned the company.
- A review of 57 neural-tech companies found only six explicitly said they would not sell users’ brain data; more than two-thirds allowed data to pass to a new owner after an acquisition.
- Studies cited in the article indicate EEG recordings may reveal identifying or sensitive information, but they do not establish what can be inferred from Muse recordings specifically.
- The article advises users to check whether their device maker uses EEG data for AI training and offers opt-out or deletion options.
Article Details
- Event Type
- Consumer data privacy and consent issue involving EEG data used for AI training
- Impact
- Interaxon used about 50,000 hours of users’ EEG recordings to train a brain foundation model without explicitly informing them. The company removed names and email addresses, but research cited in the article raises concerns that EEG data may still enable re-identification or sensitive inferences. After Science questioned the company, Interaxon updated its privacy policy to say de-identified data is used to train machine learning models.
