HugoScore hugoscore.org

Full review

Inciteful Med CAIHL draft report

Evidence-linked HugoScore draft report for a health AI tool that affects patients.

Agency checklist · October 30, 2026

Leans expands 9 of 12 yes

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

Assessed function: Inciteful Med patient-initiated medical evidence retrieval and synthesis (PubMed-grounded reports from the patient's questions and records)

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
provides a SaaS-based platform that uses Artificial Intelligence Source

Terms section 2 'Artificial Intelligence'.

2. Operator and model maker namedNonot-documented
Not independently replicated
Quotation not recorded; evidence repair pending. Source

Inciteful, Inc. is named. Terms and privacy refer to 'third party large language models' and 'AI providers' without naming them. Help pages name ChatGPT and Claude only as comparisons.

3. Patient chooses to use itYesdocumented
Not independently replicated
You may choose to provide us with medical information when you use the Platform. Source

Patient signs up, asks questions and chooses what records to add.

4. Patient sets the taskYesdocumented
Not independently replicated
Upload your records, lab results, or simply ask a health question. Source

Patient sets the question. Targeted advertising is a separate concern (question 12).

5. Patient can decline or stopYesdocumented
Not independently replicated
You may choose to provide us with medical information when you use the Platform. Source

Self-started rule applied (cites question 3 source).

6. Patient sees the AI outputYesdocumented
Not independently replicated
creating the results You receive Source

Patient sees the report and can save it as a PDF or copy it.

7. Patient can correct or challengeYesdocumented
Not independently replicated
Contact us and we'll take a look. Include the thread URL Source

Weak yes. Help article also advises follow-up questions and editing the patient profile, but no route specific to correcting an error in an output is documented.

8. Patient can get a copyYesdocumented
Not independently replicated
You can download records or delete them. Source

Help also documents a PDF report download and a Markdown copy button.

9. Patient can deleteYesdocumented
Not independently replicated
You can delete individual records, individual patients, or your whole account at any time. Source

Also How do I delete my data? help article.

10. Training or sale does not override patient choiceYesdocumented
Not independently replicated
We do not use your records, questions, or reports to train models - ours or anyone else's. Source

Source is a help-center page (last reviewed 2026-10-22), not a binding document. Terms section 15 grants a license to use Input 'for the purposes of providing or improving the Services generally' and section 16 allows de-identified data use 'for any lawful purpose'. Privacy says third-party AI providers 'contractually promise data privacy'. Flip to no if only binding documents count.

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

Vendor-hosted only.

12. No advertising or data businessNodocumented
Not independently replicated
to serve You with targeted advertising Source

Terms section 12 (AdTech): data shared with 'third-party advertising and analytics partners'. Privacy: marketing and 'advertising campaigns'. Help page says records are not shared with advertisers; that does not cover AdTech 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 report

Inciteful Med: CAIHL reassessment

September 8, 2026. Published AI-assisted draft.

Inciteful Med offers patient-accessible research with source excerpts that can support checking and better questions. Personal-data correction is distinct from correcting a synthesis. Data-use conditions and independent validation remain qualified.

Scope and agency posture

Patient-initiated evidence retrieval and synthesis, with 84 and potentially agency-expanding posture preserved from August 9.

Posture: Potentially agency-expanding.

Axis: 84/100, retained provisionally. No numerical recalibration was performed.

Sources and documented findings

  • The current method page describes question expansion, current PubMed retrieval, ranking and a citation with source excerpt for each sentence. This is a concrete checkability design claim, not proof that citations support every generated claim.
  • Current terms retain permission to create aggregate health and de-identified data for any lawful purpose and instruct users not to submit identifying information. The privacy policy says unnamed enterprise LLM providers may see inputs, recommends removing identifiers, and documents in-platform personal-information update and deletion plus a contact route. Those are data rights, not necessarily synthesis-correction rights.
  • The August report's homepage no-training promise, physician-testing statistics and absence of pre-delivery doctor review remain prior evidence. This pass did not independently reproduce those claims or validate report accuracy.

Patient authority

Inference from the documented workflow: Patients set the research question and can use evidence in their own decision process. The platform chooses expansion, retrieval and ranking, so a patient still needs to examine whether the generated framing matches their concern.

Critical capacity

Editorial assessment: Preserved excerpts and papers make disagreement and verification possible. This is the strongest rationale for the existing posture, rather than free access or founder intent. Checkable citations should not be described as guaranteed correct synthesis.

Informed control

Editorial assessment: Input minimization falls partly on users, and unnamed processors and derivative-data permissions constrain informed participation. Personal-data amendment is documented, but it must not be represented as a guaranteed correction of a report.

Assessment and limits

Keep the existing qualified posture. The current evidence method and legal conditions substantiate both its agency promise and limits. No new fact justifies a score change.

Confidence: Medium for documented method and conditions.

Remaining uncertainty: Citation entailment, ranking bias, report-correction workflow, processor identity and retention were not independently tested.

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.