Health Data Partnerships and Privacy Gaps: Ten Healthtech Firms to Watch — Research Library

Key Highlights

Google Health announced on 12 October 2026 a multi-year partnership with Flo Health, a period-tracking application reporting more than 80 million monthly users, alongside an undisclosed strategic investment.
Days later the claims window closed in a $59.5 million settlement over allegations of non-consensual data sharing, with Google to contribute $48 million pending a federal judge's final approval on 29 October.
Ten smaller health technology firms are being assessed for clinical AI that pairs diagnostic accuracy in under-resourced settings with de-identification by design.

Context and Core Dynamics

From 2027, premium subscribers may link cycle data to readings from Google wearables. Three days after the announcement, the claims window closed in the settlement; allegations stated that the application allowed Google, analytics provider Flurry and Meta to record period, ovulation and pregnancy data without consent between 2016 and 2019. Google states that participation will be opt-in, that consent may be withdrawn at any time, and that data will be neither sold nor used for advertising. The episode exposes a structural asymmetry: consumer health applications commonly sit outside HIPAA, which governs records held by institutions such as hospitals and insurers. The Federal Trade Commission's action against online therapy platform BetterHelp, which agreed to pay $7.8 million after allegations that it shared users' email addresses and mental health questionnaire answers with Facebook, Snapchat and other advertisers despite privacy assurances, follows the same pattern. The ten companies under review were each built around a defined clinical requirement and an explicit privacy position. AI Diagnostics of South Africa holds SAHPRA approval for Ostium, a digital stethoscope that applies a model trained on more than 50,000 lung recordings to flag likely tuberculosis in under five minutes, enabling community health workers to screen without radiography. Ourself applies AI research partner Stella to hormonal health, storing identity and health data separately with the linking key held on the user's device.

Caregiving Application and Social Impact

Tuberculosis remains the leading infectious cause of death worldwide, with an estimated 1.23 million deaths in 2024, and symptom checklists used under WHO guidance achieve roughly 71 per cent sensitivity, so close to three in ten cases are missed. A five-minute, radiography-free triage step is materially relevant to home-care and residential-care settings without imaging, and to rehabilitation and chronic respiratory care. Hormonal tracking tools support monitoring that often falls to unpaid family carers.

Ethics and Compliance Note

Sensitive health data such as symptoms, diagnoses, medications, mood and sexual history carries consequences beyond inconvenience, particularly where abortion is criminalised or records reach an employer. Consent architecture, de-identification, purpose limitation and data minimisation are the operative safeguards; the Flo and BetterHelp matters indicate that contractual assurances alone are insufficient. Device-linked inference warrants treatment as a distinct risk category.