HugoScore hugoscore.org

Full review

Alcott CAIHL draft report

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

Back to profile Markdown source

HugoScore CAIHL Draft Report: Alcott

  • Status: Draft for human review
  • Last reviewed: 2026-07-01
  • Review method: Public-source review of Alcott homepage, patient, family, clinic, how-it-works, pricing, privacy, terms, integrations, one-pager, and visit app surfaces; public HTML/app-shell inspection of visit.alcottai.com. No hands-on authenticated app walkthrough, vendor interview, patient interview, clinic deployment review, network-traffic inspection, security audit, or independent outcome evaluation was performed.
  • Service: Alcott
  • Vendor: Alcott, Inc.
  • URL: https://alcottai.com/
  • Category: Patient care navigation and health copilot AI

Summary

Alcott is a hybrid visit companion for patients, families, and clinics. Public materials describe a before/during/after visit workflow: patients or caregivers share symptoms, history, medications, questions, and priorities before a visit; Alcott generates patient-facing agendas and clinician-facing briefs; patients can record a visit; the system transcribes and summarizes the encounter; and summaries, follow-up tasks, reminders, documents, and family updates can be shared after the visit.

From a CAIHL perspective, Alcott is unusually patient-agency-forward in its public framing. It says patient and family access is free, emphasizes patient data control, supports family/caregiver visibility, and offers practical action support around remembering, preparing, sharing, and following through. The terms and privacy policy are also more detailed than many early health AI products: they disclose audio deletion after transcription, named subprocessors and integrations, health-content retention rules, access/download/delete rights, and a provider Business Associate Agreement.

The main caution is that the tool is not purely patient-controlled. Clinics pay for Alcott's coordination layer, receive pre-visit intelligence briefs, can connect CRM/EHR workflows, and can receive generated summaries or completed forms. That hybrid model may expand patient agency when patients control sharing and correction, but may constrain agency if patient-reported material moves into institutional workflows without strong review, correction, refusal, and revocation controls.

  • Agency posture: Potentially agency-expanding, mixed
  • Agency axis position: 72 of 100
  • Confidence: Medium-low draft, official sources only

Evidence Reviewed

CAIHL Profile

Who does this AI serve?

Alcott serves patients and families directly, but also serves clinics that pay for pre-visit preparation, intake automation, EHR/CRM delivery, and workflow relief. The public model is therefore hybrid: patient-leaning in user experience and framing, but clinic-integrated in deployment and economics.

Can patients tell AI is involved?

Yes. The public pages describe AI-guided visit preparation, AI-powered visit summaries, scribe transcription, chat over uploaded documents, and clinician-facing intelligence briefs. Patient visibility is less certain downstream: public materials do not fully show how patients see or approve the final clinician brief, completed form, or EHR write-back.

Can patients meaningfully choose?

Partially documented. Self-directed patient and family use appears voluntary and free. The visit app says recording requires a signed-in account for privacy and asks users to confirm clinician agreement before recording. Clinic-invited prep links, SMS reminders, family/caregiver access, and provider-integrated workflows need hands-on review to confirm opt-out, refusal, and proxy-consent mechanics.

Can patients correct or challenge what the AI produces?

Partially documented. The privacy policy says users can access, download, delete information, and close their account. App text suggests notes can be added to sections and share links can expire or be revoked. Public materials do not yet clearly document patient correction of AI-generated clinician briefs, completed intake forms, EHR write-back, transcripts, or visit summaries once shared with a provider.

Does it help patients understand or act?

Yes. Alcott's strongest agency claim is action support: visit agendas, summaries, action items, reminders, document sharing, family updates, calendar/task export, and clinician handoff materials. This is closely aligned with the CAIHL praxis cycle of reflection, strategic action, and continuity of patient narrative.

Agency Interpretation

Alcott is promising because it gives patients a concrete way to prepare their story, remember what happened, involve trusted caregivers, and carry visit context across encounters. The patient/family-free model avoids putting a paywall directly on understanding one's own care.

The agency risk comes from mediation. Alcott can move patient-reported information into clinician-facing briefs, forms, CRM objects, EHR-connected workflows, and family-sharing links. Those flows can support patients, but only if patients can see what was generated, correct errors, revoke access, understand who saw what, and refuse use without losing care access.

Key Unknowns

  • Actual onboarding, consent, sharing, deletion, export, and correction UX.
  • Whether patients can review and edit clinician-facing briefs before provider delivery.
  • How transcript or summary errors are corrected after sharing.
  • How completed intake forms are previewed, corrected, retained, and delivered.
  • Whether EHR/CRM write-back can be contested or amended by the patient.
  • How clinic-invited prep links handle refusal, expired links, identity verification, and alternate non-AI workflows.
  • Whether family/caregiver access has granular consent, revocation, audit, and conflict safeguards.
  • Whether analytics and advertising pixels are fully isolated from PHI and authenticated app activity.
  • Whether AWS Bedrock, AWS Transcribe Medical, Supabase, Lambda, S3, Keragon, HubSpot, Practice Better, Twilio, Stripe, and Google flows match the privacy policy in production.
  • Whether Alcott's early pilot metrics are independently validated.
  • Accessibility, multilingual accuracy, low-literacy usability, disability support, interpreter workflows, and caregiver/proxy safety.
  • Regulatory posture beyond Alcott's statement that the service is not a medical device.

Publication Recommendation

Ready for human review as a source-backed draft profile. Publish as a draft, not a reviewed or verified profile. Keep the confidence at medium-low until hands-on app review, vendor clarification, independent evidence, analytics isolation, correction/contestability workflows, family-sharing governance, and EHR write-back behavior are verified.