Full review
VYTL CAIHL draft report
Evidence-linked HugoScore draft report for a health AI tool that affects patients.
Agency checklist · October 30, 2026
Leans expands 8 of 12 yes
Document-based draft. Independent confirmation is incomplete. Hugo has not personally verified this service.
Assessed function: Historical patient-initiated fax collection and AI record organization, with separate clinic accounts. The August 9 baseline is 84, potentially agency-expanding, mixed.
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 Service uses automated tools and artificial-intelligence systems, including third-party AI servicesSource Terms section 5 and Privacy section 4 say AI reads and extracts documents. The terms are accepted at account creation. The home page does not mention AI. |
| 2. Operator and model maker named | No | not-documented Not independently replicated | Quotation not recorded; evidence repair pending. Operator is named (VYTL Health LLC, founder Dr. Kinaan Javed on the About page) but the AI is described only as 'third-party AI services'. No model maker is named. |
| 3. Patient chooses to use it | Yes | documented Not independently replicated | Request your records from any provider, and we organize everything in one secure place.Source Patient requests records and has them faxed to a personal VYTL number. Clinic accounts exist separately and were not assessed. |
| 4. Patient sets the task | No | not-documented Not independently replicated | Quotation not recorded; evidence repair pending. The patient picks which records to request or upload, but the AI step (read, classify, extract) is fixed by the operator and no patient control over it is documented. |
| 5. Patient can decline or stop | Yes | documented Not independently replicated | Request your records from any provider, and we organize everything in one secure place.Source Self-started tool rule: Q3 is yes and no care depends on it. Source is the Q3 source. |
| 6. Patient sees the AI output | Yes | documented Not independently replicated | You are responsible for reviewing extracted information against your original recordsSource Privacy section 4. The patient sees the extracted data and the original document. |
| 7. Patient can correct or challenge | Yes | documented Not independently replicated | You can correct, export, or delete your records at any time by emailingSource Also in Privacy section 8 (through account settings or [email protected]). Terms say the original upload, not the extracted data, is authoritative. |
| 8. Patient can get a copy | Yes | documented Not independently replicated | You may access, correct, export, or delete your personal information through your account settingsSource Privacy section 8. |
| 9. Patient can delete | Yes | documented Not independently replicated | Deletion is permanent within thirty days; only the access log required by law is retained.Source Terms section 12 describes a 30-day grace period before permanent erasure, and compliance records are kept. |
| 10. Training or sale does not override patient choice | No | not-documented Not independently replicated | Quotation not recorded; evidence repair pending. No statement about model training or an opt-out. Privacy section 2 lists 'Improve, develop, and analyze the Service' as a use. Subprocessors may use data only to provide services to VYTL. 'We do not sell your personal information' covers sale only. |
| 11. Patient can run it or take it elsewhere | No | not-documented Not independently replicated | Quotation not recorded; evidence repair pending. Vendor-hosted. Export is by account settings or email only, and no way to run it yourself or move it is documented. |
| 12. No advertising or data business | Yes | documented Not independently replicated | We do not sell your personal information.Source Section 3 lists sharing only with providers you authorize, service providers, legal needs and business transfers. De-identified data, licensing and ads are not mentioned either way. Revenue is subscriptions and clinic plans. |
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 report
VYTL: CAIHL reassessment
September 8, 2026. Published AI-assisted draft.
VYTL’s historical profile describes patient-directed record collection and administrative support. Current primary sources could not be recovered sufficiently to refresh the judgment. Its 84 is an unrefreshed baseline, with current verification incomplete.
Scope and agency posture
Historical patient-initiated fax collection and AI record organization, with separate clinic accounts. The August 9 baseline is 84, potentially agency-expanding, mixed.
Posture: Potentially agency-expanding, mixed.
Axis: 84/100, retained provisionally. No numerical recalibration was performed.
Sources and documented findings
- Current direct attempts to read the homepage, privacy policy and terms failed in the public retrieval tool. One targeted domain search returned no results. This is a retrieval limitation, not proof that the service has closed or lacks policies.
- The frozen August report documents patient-directed record requests and sharing, authoritative originals, access/correction/export rights and vendor-hosted processing through unnamed AI services. Those are historical source findings and were not refreshed successfully.
Patient authority
Inference from the documented workflow: The historical workflow gives patients a way to gather and direct their records. Whether they can revoke shares, preserve a usable copy independently or correct classifications is still unestablished, and should not be inferred from the word ownership.
Critical capacity
Editorial assessment: Access to originals could let patients check extracted dates and labels. The historical administrative scope does not claim medical interpretation. Current availability and the actual comparison/correction workflow remain unknown.
Informed control
Editorial assessment: Processor identity, training conditions and export details cannot be resolved from inaccessible current materials. A failed lookup warrants lower confidence in freshness, not a numerical penalty or an accusation about the vendor.
Assessment and limits
Do not silently carry the August assessment forward as newly verified. Keep the baseline for audit continuity and flag the current deployment and conditions as unclear.
Confidence: Low for current evidence, historical baseline only.
Remaining uncertainty: Current operation, policy text, sharing revocation, extraction correction, export and pricing were not established.
Published documentation is evidence of stated conditions, not proof of actual implementation. Unknowns did not receive automatic negative points. Funding, sponsorship, and public code do not determine agency by themselves.
Review provenance
- Reviewer/model: OpenAI Codex / GPT-6.
- 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.
- 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.
- Earlier review: 2026-08-09. Historical assessment. Earlier claims are not automatically reverified by this publication.