HugoScore hugoscore.org

Condition-specific coaching and care AI

Virta Health

Virta distinguishes optional patient-facing Viv conversations from background care algorithms. Members are offered a choice of Viv or a human coach. That choice does not establish the ability to refuse all AI processing or change the program’s objectives.

Document-based draft Read full report Open directory

Checklist assessment dated October 30, 2026. Independent confirmation is incomplete. Hugo has not personally verified this service.

Leans constrains3 of 12 agency checks metDocument-based draft. Independent confirmation incomplete; not product-tested.How categories work

Agency checklist · October 30, 2026

Leans constrains 3 of 12 yes

Document-based draft. Independent confirmation is incomplete. Hugo has not personally verified this service.

Assessed function: Virta Health human-led metabolic program with care-team algorithms and an optional patient-facing AI assistant (Viv) and Virta Metabolic Intelligence. Different AI functions have different degrees of patient control.

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
QuestionAnswerEvidence / reviewSource and reasoning
1. Told it is AIYesdocumented
Not independently replicated
The Service incorporates artificial intelligence ("AI"), machine learning, and other automated technologies as part of delivering care Source

Terms Section 7 states AI use and has users acknowledge and consent by using the Service. Viv is labeled 'AI Assistant' in press and the app listing.

2. Operator and model maker namedNonot-documented
Not independently replicated
We may use third-party AI service providers to deliver features of the Service. Source

Virta Health is named. The AI model makers are not named, and Virta 'may select, change, or replace these providers'.

3. Patient chooses to use itNodocumented
Not independently replicated
We use AI and automated technologies to support clinical decision-making, provide personalized health recommendations and guidance Source

Care-team algorithms are applied within the program. For Viv alone, the press release says 'Members can choose whether they’d like to contact Viv', so Q3 would be yes for Viv.

4. Patient sets the taskNodocumented
Not independently replicated
offers in-the-moment, personalized support for common, non-clinical requests Source

Viv covers a scoped set of non-clinical requests. Members ask the questions, but the program defines the task. Care-team algorithms are not patient-directed.

5. Patient can decline or stopNonot-documented
Not independently replicated
Members can choose whether they’d like to contact Viv for specific questions Source

Viv is optional with the health coach as an alternative. A way to decline the care-team algorithms and keep the same care is not documented.

6. Patient sees the AI outputNonot-documented
Not independently replicated
Quotation not recorded; evidence repair pending. Source

Members see Metabolic Intelligence insights and Viv responses (app update: 'ask Viv about menus, nutrition labels'), but outputs of care-team algorithms are not shown to members in the sources.

7. Patient can correct or challengeNonot-documented
Not independently replicated
Quotation not recorded; evidence repair pending. Source

No documented route to correct or contest AI output. Providers 'maintain oversight'.

8. Patient can get a copyYesdocumented
Not independently replicated
Right to Know/Access: Confirm whether we process your personal data and obtain a copy in a portable format. Source

Privacy policy state-law rights. The HIPAA Notice also gives a right to an electronic copy of the medical record, usually within 30 days.

9. Patient can deleteYesdocumented
Not independently replicated
Request Deletion: You may request that we delete your personal information. Source

Subject to legal retention.

10. Training or sale does not override patient choiceNodocumented
Not independently replicated
We may use de-identified and aggregated data derived from your use of the Service to train, develop, and improve our own internal AI models. Source

Terms 7(c). Third-party AI providers are barred from using data for their own training (7(b)), but Virta's own training use has no patient opt-out documented.

11. Patient can run it or take it elsewhereNonot-documented
Not independently replicated
Quotation not recorded; evidence repair pending. Source

Vendor-hosted. No patient route to run it or move the function.

12. No advertising or data businessNodocumented
Not independently replicated
Business Partners: De-identified data may be shared for research or business purposes. Source

Privacy policy also documents interest-based advertising and advertising device IDs in the apps. The HIPAA notice says PHI is not used in AI for marketing or sale. Privacy policy says Virta does not sell personal data.

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.

MixedEarlier review, before the published method. Documents only.

Agency finding, before method v1.3

Mixed, potentially agency-expanding

Legacy editorial placement, pre-method: 54

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.
The question we ask Who does Virta Health serve in this deployment?
Control Virta Health human-led metabolic program with care-team algorithms and an optional patient-facing AI assistant (Viv) and Virta Metabolic Intelligence. Different AI functions have different degrees of patient control.
Agency read Leans constrains
Vendor
Virta Health
Who it serves
Hybrid patient-facing metabolic care with mostly behind-the-scenes AI
Primary User
Members with obesity, prediabetes, diabetes, coaches, providers, employers, and health plans
Control Model
Professional or institution-configured service. Patient controls depend on the deployment.
Patient Impact
On-demand nonclinical guidance and human contact support practical decisions. Program-level clinical results are not evidence that AI explanations or flags are inspectable.
Profile Status
Document-based draft
Last assessed
Oct 30, 2026
Earlier review confidence
Moderate for documented functions and policy (AI-assisted draft)
Earlier review model
OpenAI Codex / GPT-6
Earlier human review record
Hugo Campos authorized publication of these AI-assisted draft reassessments on September 8, 2026. This does not claim comprehensive human verification of every finding.

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: 54

Method v1.3. The control line is evidence for how the agency finding arose, not a quality score. Legacy editorial placement, pre-method: 54. Not comparable with v1.3 positions.

Mixed, potentially agency-expanding

Virta distinguishes optional patient-facing Viv conversations from background care algorithms. Members are offered a choice of Viv or a human coach. That choice does not establish the ability to refuse all AI processing or change the program’s objectives.

Patient agency

How this tool changes agency

Expands agency when

On-demand nonclinical guidance and human contact support practical decisions. Program-level clinical results are not evidence that AI explanations or flags are inspectable.

Limits agency when

Current Viv availability, model-output correction and practical requests to restrict AI remain untested

Patient agency assessment

Who sets and changes the goal?

Patient authority

Choice between Viv and a coach is documented. That choice does not mean members can opt out of all background algorithms or change the program's clinical objectives.

What can the patient understand, question, or do?

Critical capacity

On-demand nonclinical guidance and human contact support practical decisions. Program-level clinical results are not evidence that AI explanations or flags are inspectable.

Can the patient evaluate the conditions of use?

Informed control

The new public notice provides substantive AI information missing from the seed. Formal records rights and limits on AI restriction are more informative than app-store deletion labels alone.

Text findings

Conditions of use

Published controls and their limits

The new public notice provides substantive AI information missing from the seed. Formal records rights and limits on AI restriction are more informative than app-store deletion labels alone.

What remains unknown?

Not tested or not established

Current Viv availability, model-output correction and practical requests to restrict AI remain untested

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

Current Viv availability, model-output correction and practical requests to restrict AI remain untested 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 documented functions and policy (AI-assisted draft)