Patient care navigation and health copilot AI
My Doctor Friend
My Doctor Friend is a patient and caregiver health companion from My Doctor Companion, Inc., publicly authored by physician-founder Michael Turken, MD, MPH and positioned as "not an AI doctor, but your thought partner in health." The refreshed draft adds the Working Health Record consolidation shipped June 2026, health record connections, metered Visit Companion sessions, Apple Health and voice features, and published pricing from a limited free tier to Pro at $100 per month. It flags that the privacy notice changed June 20, 2026 and could not be re-read in this pass, that App Store labels link health, location, financial, and sensitive data to identity, that the free-tier allowance is described inconsistently, and that the prior review's OpenAI/Anthropic-under-BAA findings need re-verification.
Public-source research has been drafted, often with AI assistance. No comprehensive human review is recorded unless the profile provenance says so.
Summary judgment · 82% toward patient-directed
Potentially agency-expanding
Explicitly patient- and caregiver-directed, with a physician-founder, thought-partner framing, and record consolidation moving toward longitudinal patient custody. The posture stays qualified because the June 20, 2026 privacy notice is unread since its update, audio and Apple Health handling are unverified, correction workflows are undocumented, and the deepest agency features are metered by $20-100 monthly tiers.
Patient agency
How this tool changes agency
Working Health Record consolidation, symptom tracking, templated doctor-ready summaries, in-visit question support, family profiles, and suggested questions are directly action-oriented.
The prior privacy notice documented data correction and portability rights, but app-level editing of the Working Health Record, generated summaries, and condition suggestions remains undocumented, and the June 20, 2026 notice needs re-reading.
Patient-facing signals
Who does this AI serve?
The service is built for patients, family caregivers, and care preparation rather than clinician documentation.
Can patients tell AI is involved?
AI Q&A, voice, visit companion, and summary generation are visible product features.
Can patients meaningfully choose?
Use is voluntary and a free tier exists with voice included, but unlimited messages, Visit Companion sessions, and health record connections require Plus ($20/mo), Family ($35/mo), or Pro ($100/mo) subscriptions.
Can patients correct or challenge what the AI produces?
The prior privacy notice documented data correction and portability rights, but app-level editing of the Working Health Record, generated summaries, and condition suggestions remains undocumented, and the June 20, 2026 notice needs re-reading.
Does it help patients understand or act?
Working Health Record consolidation, symptom tracking, templated doctor-ready summaries, in-visit question support, family profiles, and suggested questions are directly action-oriented.
Text findings
Who is left out or burdened?
Cost, platform, and language boundaries now explicit
iPhone-only (iOS 16+), English-only, rated 16+, with the deepest agency features behind $20-100 monthly tiers. Accessibility conformance, low-literacy support, and older-adult usability remain undocumented; the Family tier supports multi-profile household caregiving.
What happens to patient data?
Previously well-documented, currently unverified
The prior notice recorded OpenAI and Anthropic processing under BAAs, retention, deletion, and correction/portability rights. The notice was modified June 20, 2026 and could not be machine-read in this pass, while App Store labels link health and fitness, location, financial, sensitive, and user-content data to identity. Re-verification is the profile's top follow-up.
Are the clinical boundaries clear?
Clear in wording, high-stakes in use
"Not an AI doctor" framing, no-diagnosis/no-prescription disclaimers, and 911 emergency direction are consistent across surfaces. Condition matching, get-care guidance, and confident conversational output still need safety validation.
Who defined what good looks like?
Physician-founder and vendor-defined
Michael Turken, MD, MPH is publicly identified as founder, which adds accountability, but no independent clinical safety, accuracy, bias, accessibility, or patient-outcome evaluation was found, and the 5.0 App Store average rests on 39 ratings.
Review provenance
Criteria
Same HugoScore CAIHL-derived criteria for every tool; public criteria are displayed from site/data/criteria.json, with fuller method notes in SCORING_FRAMEWORK.md.
Reviewer
AI-assisted public-source refresh prepared in Claude (Cowork); no named human reviewer recorded.
AI / model
Claude Fable 5 (claude-fable-5).
Human review
No comprehensive human review has been completed or claimed for this draft profile.
Review date
2026-08-09
Limitations
Privacy notice and terms not machine-readable in this pass (client-side rendering); prior-review findings carried forward where marked. No vendor interview, hands-on testing, audio-handling verification, security audit, accessibility audit, or independent validation.
Review method
Refreshed public-source review of the marketing site and pricing page, the US App Store listing including version history and privacy labels, and public founder context. The privacy notice and terms could not be machine-read in this pass (client-side rendering); 2026-06-08 findings are carried forward where noted and flagged for re-verification against the June 20, 2026 privacy-notice update. No vendor interview, hands-on testing, audio-handling verification, security audit, accessibility audit, or independent validation.
Draft profile · Medium draft, official sources only; privacy-notice re-read pending