Samsung says withdrawing consent for AI training in Samsung Health will not delete users’ existing health records or stop them from using the app. The clarification followed an in-app warning that appeared to say opting out would disable account syncing and erase stored health information.
Only data collected separately for AI development will be removed, Samsung said. The warning instead raises questions about how clearly the company explains its use of medical, reproductive, medication, and fitness information for AI development.
A warning that appeared to threaten health histories
The consent screen, titled “Consent to the Use of Health Data for AI Training and Modelling,” asks users to let Samsung use information from Samsung Health to improve health-analysis algorithms and AI features. It also says the process may include human review.
The listed information includes body measurements, nutrition, step counts, activity, and sleep. It also covers prescriptions, diagnoses, test results, treatment records, and menstrual-cycle information, according to details reviewed by Android Authority.
Those categories show how smartwatch health tools can build extensive personal records from cardiovascular, sleep, activity, and other signals. Samsung’s push into AI-powered health insights also raises broader questions about how companies turn sensitive wearable data into advice, particularly as Apple and Samsung compete to interpret glucose and other health signals.
The controversy began with the withdrawal message. It appeared to say that users who revoked consent could no longer sync data to their Samsung accounts and that stored health information would be deleted unless Samsung had to retain it by law.
Samsung later said the message referred only to the separate AI-development dataset. In a statement provided to 9to5Google, the company said existing records would remain available and Samsung Health would continue working.
Testing conducted by SamMobile and cited by Android Authority found that records remained visible and syncing continued after consent was withdrawn. Samsung said it was revising the notice to make its meaning clearer.
Samsung has not published a complete list of affected regions. It also has not explained whether information involved in human review is identifiable, pseudonymized, or aggregated.
The consent wording creates an IT policy problem
The prompt can affect organizations that permit Samsung Health on managed devices, maintain bring-your-own-device policies, or recommend Samsung wearables through employee wellness programs.
Administrators should not tell users that withdrawing AI consent will erase their Samsung Health histories or permanently stop syncing. Internal guidance should separate information required to operate Samsung Health from data copied for AI development.
Privacy assessments should document which information is covered, whether human review is involved, how long AI-development data is retained, and whether the notice differs by region. Reviews of AI tools connected to sensitive systems should also examine permissions, access controls, logging, and audit records.
Samsung’s US consumer health data privacy statement covers medical records, diagnoses, medication, reproductive information, and data produced through algorithms or machine learning. It also says consumers may withdraw consent where consent is the applicable legal basis.
That broader policy does not explain how the new AI-training process handles data separation, human review, retention, or regional availability. Samsung has clarified what the opt-out deletes, but the consent screen still leaves those operational details unresolved.
Read more: Organizations updating wearable policies should also account for how smart glasses are reshaping privacy and device rules in courts and other sensitive environments.