Condition-specific coaching and care AI
Twin Health
Twin Health combines sensor-based guidance and clinical care with employer and health-plan contracts. Patient authority when personal goals differ from program cost or medication-reduction goals remains unclear. Current policy documents correction rights and a separate clinical-data notice.
Checklist assessment dated October 30, 2026. Independent confirmation is incomplete. Hugo has not personally verified this service.
Agency checklist · October 30, 2026
Mixed 5 of 12 yes
Document-based draft. Independent confirmation is incomplete. Hugo has not personally verified this service.
Assessed function: Twin Health AI digital-twin metabolic care program with a clinical team, offered through employer and health plan contracts. Not a standalone patient chatbot.
A no can mean missing documentation or institution-dependent control; it does not by itself establish harmful behavior. These categories are not clinical safety or product-quality ratings.
- Public documents do not establish actual product behavior.
- The independent review screened rules and selected evidence; it did not replicate all twelve answers for this tool.
See all twelve answers and their evidence
| Question | Answer | Evidence / review | Source and reasoning |
|---|---|---|---|
| 1. Told it is AI | Yes | documented Not independently replicated | The AI digital twin™ shows your metabolism in real-time using data from smart devicesSource The member page and privacy policy call it an 'AI digital twin' and say Twin may use AI technology to process personal information. In-app wording and the telehealth informed consent were not viewed. |
| 2. Operator and model maker named | Yes | documented Not independently replicated | Our AI digital twin™ technology and clinical care team work togetherSource Twin Health, Inc. is named as operator and presents the digital twin as its own technology. The model behind its 'generative AI tool' is not named. |
| 3. Patient chooses to use it | No | deployer-dependent Not independently replicated | A fully covered program through your organizationSource Access is through an employer or plan sponsor and eligibility is set through the sponsor. No source states enrollment is voluntary, although registration requires patient action (Wired describes a member choosing to 'give it a try' when an employer offered it). |
| 4. Patient sets the task | No | documented Not independently replicated | Twin gets paid only when users achieve certain clinical outcomes, such as lower blood sugar, weight loss, or the reduction of metabolic medications.Source Operator and sponsor define the clinical goals and Twin is paid on outcomes. Members choose among recommendations, but the purpose is set by the program. |
| 5. Patient can decline or stop | No | not-documented Not independently replicated | Quotation not recorded; evidence repair pending. Source No documented way to decline or stop the AI and keep the same care. Terms say Twin may terminate access, including if a member becomes inactive. |
| 6. Patient sees the AI output | Yes | documented Not independently replicated | You can see it in your app once you become a member.Source The member FAQ says the Digital Twin is visible in the app. |
| 7. Patient can correct or challenge | No | not-documented Not independently replicated | Quotation not recorded; evidence repair pending. Source Members can notify Twin of errors in personal information, but no route to contest AI recommendations is documented. |
| 8. Patient can get a copy | Yes | documented Not independently replicated | You have the right to access your medical records.Source Terms also give 90 days after termination to retrieve Your Data by emailing [email protected] (commercially reasonable efforts, possible fees). |
| 9. Patient can delete | Yes | documented Not independently replicated | You may also request for your profile in the App to be deleted.Source Subject to legal retention. Some information may remain after account deletion. |
| 10. Training or sale does not override patient choice | No | documented Not independently replicated | Twin Health may use such de-identified data in any lawful manner including for our generative AI toolSource Privacy policy: normative database of de-identified data used for benchmarking, analysis and its generative AI tool. No patient opt-out documented. |
| 11. Patient can run it or take it elsewhere | No | not-documented Not independently replicated | Quotation not recorded; evidence repair pending. Source Vendor-hosted and sponsor-contracted. No patient route to run it or move the function. |
| 12. No advertising or data business | No | documented Not independently replicated | We may further share such aggregate, non-identifiable information with business partners, sponsors and other third parties.Source The policy also says Twin will not sell personal information and that no information shared within the Services is shared with third-party advertisers. Wired reports employers receive aggregated, anonymized outcome reports. |
Initial coding: Claude Sonnet helpers, compiled by Claude Opus 5.5 (Anthropic). Selective review and corrections: GPT-6 Astra (OpenAI). Editor: Hugo Campos. Acceptance of corrections is not product verification.
Checklist rules · Independent review and limitations · Later evidence follow-up
Earlier v1.3 assessment and control finding
Historical assessment under different rules. It does not set the current checklist zone.
Agency finding, before method v1.3
Mixed, potentially agency-expanding
Legacy editorial placement, pre-method: 58
This profile has not yet been reviewed under the published method. Its finding and any number are editorial placements made before the method existed.
- Evidence
- Documentary review. Public sources only, no hands-on testing by HugoScore.
- Review status
- AI draft. No person has settled the findings.
Earlier evidence and assessment
The sections below preserve the earlier review. Its classifications and provenance are historical; the checklist above gives the current draft classification.
Earlier finding · Legacy editorial placement, pre-method: 58
Method v1.3. The control line is evidence for how the agency finding arose, not a quality score. Legacy editorial placement, pre-method: 58. Not comparable with v1.3 positions.
Mixed, potentially agency-expanding
Twin Health combines sensor-based guidance and clinical care with employer and health-plan contracts. Patient authority when personal goals differ from program cost or medication-reduction goals remains unclear. Current policy documents correction rights and a separate clinical-data notice.
Patient agency
How this tool changes agency
Personal data and daily feedback can support reflection. Clinical outcome studies do not establish the ability to inspect a model recommendation or negotiate its objective.
Individual sponsor reporting, clinical notice details, recommendation explanations and care-plan negotiation remain unverified
Patient agency assessment
Who sets and changes the goal?
Members receive individualized guidance, but the public pages do not establish how they can retain a preferred treatment goal when it differs from program cost or medication-reduction targets.
What can the patient understand, question, or do?
Personal data and daily feedback can support reflection. Clinical outcome studies do not establish the ability to inspect a model recommendation or negotiate its objective.
Can the patient evaluate the conditions of use?
The current policy adds concrete correction rights and a separate clinical-data scope. It does not support a blanket claim that all identifiable health data is shared with employers.
Text findings
Conditions of use
Published controls and their limits
The current policy adds concrete correction rights and a separate clinical-data scope. It does not support a blanket claim that all identifiable health data is shared with employers.
What remains unknown?
Not tested or not established
Individual sponsor reporting, clinical notice details, recommendation explanations and care-plan negotiation remain unverified
Who evaluated this?
AI-assisted public-source draft
Vendor statements describe published conditions, not independently verified behavior. No clinical, security, accessibility, or legal validation is claimed. Earlier evidence remains dated in the report and history.
Sources checked
Source-specific findings and retrieval limitations are recorded in the full report.
Earlier review provenance
Criteria
CAIHL-derived HugoScore framework and September 7 qualitative review priorities. Draft v1.2 numerical anchors remain unadopted.
Reviewer
AI-assisted public-source reassessment prepared in OpenAI Codex.
AI / model
OpenAI Codex / GPT-6
Human review
Hugo Campos authorized publication of these AI-assisted draft reassessments on September 8, 2026. This does not claim comprehensive human verification of every finding.
Review date
2026-09-08
Limitations
Individual sponsor reporting, clinical notice details, recommendation explanations and care-plan negotiation remain unverified No live product use, patient-data upload, account creation, code audit, clinical evaluation, or independent implementation validation.
Earlier review method
Focused public-source reassessment using CAIHL: patient authority, critical capacity, and informed control. Existing evidence plus one focused primary-source pass and at most one targeted follow-up. No live product testing. Numerical scores remain provisional editorial placements, not a new calculation.
Document-based draft · Moderate for program and policy statements (AI-assisted draft)