Full review
Counterforce Health CAIHL draft report
Evidence-linked HugoScore draft report for a health AI tool that affects patients.
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: Counterforce Health free AI-drafted appeal of a patient's chosen coverage denial (individual use); clinic and revenue-cycle products not assessed
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 | Counterforce Health provides an AI-powered tool to help users draft appeals for denied health insurance claims.Source Also on homepage ('state-of-the-art AI'). |
| 2. Operator and model maker named | No | not-documented Not independently replicated | Quotation not recorded; evidence repair pending. Source Counterforce Health, Inc. is named. The privacy policy refers to 'Third-party AI providers' and Terms to 'proprietary machine learning'. No model maker is named. |
| 3. Patient chooses to use it | Yes | documented Not independently replicated | I want to appeal a denialSource Sign-up form option, plus homepage 'Start Free Appeal' and login page ('Free for individual patients'). Patient initiates. |
| 4. Patient sets the task | Yes | documented Not independently replicated | analyze the information you provide and generate draft appeal letters and supporting argumentsSource Patient supplies the denial and documents. Homepage: 'will never accept money from insurance companies'; funded by grants. Clinic and RCM products (homepage) are separate. |
| 5. Patient can decline or stop | Yes | documented Not independently replicated | I want to appeal a denialSource Self-started rule applied (cites question 3 source). Terms 11: user may stop using the Services at any time. |
| 6. Patient sees the AI output | Yes | documented Not independently replicated | generate draft appeal letters and supporting argumentsSource Homepage: 'Get Your Appeal Letter'. |
| 7. Patient can correct or challenge | No | not-documented Not independently replicated | Quotation not recorded; evidence repair pending. Source Terms say to 'carefully review' the draft before submitting. Privacy offers a right to correct personal information. No documented route to edit or contest the AI draft was found on public pages. Editing may exist in the logged-in app (not verified). |
| 8. Patient can get a copy | No | documented Not independently replicated | Depending on your location, you may have rights to access, correct, delete, or opt out of certain uses of your Personal Information.Source Request route: [email protected]. The right is location-dependent per the policy. The patient also receives the letter ('Get Your Appeal Letter'). Borderline, see caveats. |
| 9. Patient can delete | No | documented Not independently replicated | Depending on your location, you may have rights to access, correct, delete, or opt out of certain uses of your Personal Information.Source Same sentence covers deletion. Location-dependent. Borderline, see caveats. |
| 10. Training or sale does not override patient choice | No | documented Not independently replicated | Operate and improve AI models: Train and validate systems; we strive to de-identify data where feasible.Source Terms 1(b) grant a perpetual, irrevocable license to use User Content to develop and improve 'our AI Models'. Homepage FAQ says 'We never share or sell your information', which conflicts with privacy permitting third-party AI providers. Answered by binding documents. No training 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 only. |
| 12. No advertising or data business | No | documented Not independently replicated | and provide interest-based advertisingSource Privacy section 5: cookies and pixels used for interest-based advertising. Also 'We do not sell Personal Information for monetary consideration'. Terms license allows anonymize and aggregate use. |
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
Counterforce Health: CAIHL reassessment
September 8, 2026. Published AI-assisted draft.
Counterforce Health offers free assistance with a patient’s chosen coverage-denial challenge. The policy discloses secondary data use. Patient control over final wording, evidence checking, submission, and training use remains incompletely verified.
Scope and agency posture
Counterforce Health offers free assistance with a patient’s chosen coverage-denial challenge. The policy discloses secondary data use. Patient control over final wording, evidence checking, submission, and training use remains incompletely verified.
Posture: Potentially agency-expanding, with AI-training and verification caveats.
Axis: 84/100, retained provisionally. No numerical recalibration was performed.
Sources and documented findings
- The homepage offers free appeal creation and describes arguments from clinical and policy sources. Homepage, freshly retrieved September 8. Does not verify the quality of any letter.
- The same page markets staff productivity and revenue recovery. Homepage, institutional metrics. These metrics do not establish who controls an individual appeal.
- The privacy policy includes training and validation among data uses, with de-identification qualified by feasibility. It lists third-party AI processing. Privacy policy, sections 3–4, recovered September 8. Actual training datasets and processing implementation are unverified.
- Access, correction, deletion, and some opt-outs depend on location. Privacy policy, section 7. No universal, specific training opt-out is established. Data correction is not proof of appeal editing.
Patient authority
Inference from the documented workflow: the individual service is directed toward a patient's chosen denial challenge. The baseline describes drafts that patients submit, but this pass did not establish an enforced final-approval gate, broad goal redirection, or independent output-use rights. Keep correction as partial rather than upgrading it from the marketing.
Critical capacity
Editorial assessment: assistance can reduce the work required to oppose a coverage decision. It remains unclear how well patients can inspect the argument's evidence, compare alternatives, or challenge the system itself. An advocacy mission does not demonstrate those capacities.
Informed control
Editorial assessment: training disclosure makes a consequential practice visible, while the scope of patient choice over that practice remains uncertain. This supports the existing caveat. It is not evidence that every patient's information is used in training.
Assessment and limits
Counterforce Health offers free assistance with a patient’s chosen coverage-denial challenge. The policy discloses secondary data use. Patient control over final wording, evidence checking, submission, and training use remains incompletely verified.
Confidence: Medium for published documentation, low for implementation.
Remaining uncertainty: Final submission authority, evidence traceability, output-use terms, training opt-out, and the reach of deletion.
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: 2026-06-10. Historical assessment. Earlier claims are not automatically reverified by this publication.