Therabot CAIHL draft report

Therabot: CAIHL reassessment

September 8, 2026. Published AI-assisted draft.

Therabot is assessed here as a supervised research deployment. Trial participation and researcher oversight do not establish the controls of a current commercial product. Its former 45 is retained as history, outside current consumer-score comparisons.

Scope and agency posture

The supervised research deployment documented in Dartmouth's trial, not an assumed current commercial therapy product.

Posture: Research-stage, unclear in deployment.

Axis: Unscored in current comparisons. Historical axis: 45/100. Supervised research does not establish a current consumer-product score. This is a scope change, not a numerical downgrade.

Sources and documented findings

• Dartmouth describes participant-initiated conversations, researcher review and crisis escalation readiness in the 2025 trial. The comparison was a waitlist group, not a head-to-head trial against human therapists. https://home.dartmouth.edu/news/2025/03/first-therapy-chatbot-trial-yields-mental-health-benefits

• Dartmouth's April 2026 notice describes Therabot Labs licensing, funding and sublicensing work. It does not document a public product's controls or availability. https://home.dartmouth.edu/news/2026/04/brian-pogue-and-nick-jacobson-receive-innovation-awards

Patient authority

Inference from the documented workflow: Participants could initiate conversations inside a researcher-governed intervention. Enrollment restrictions are a research scope boundary, not evidence of coercion or diminished agency within participation.

Critical capacity

Editorial assessment: Symptom improvement and therapeutic alliance do not themselves demonstrate patients becoming better able to interrogate advice, evaluate evidence or redirect therapeutic goals.

Informed control

Editorial assessment: Supervision is documented for the trial. Future commercial privacy, withdrawal, correction and oversight cannot be inferred from research safeguards.

Assessment and limits

The historical 45 is a placeholder for research-stage uncertainty, not a measured agency deficit. The baseline's broad praise of an evidence standard should not substitute for assessing patient authority.

Confidence: Moderate for historical trial deployment, low for present commercialization.

Remaining uncertainty: Public commercial availability and governance are not established by the bounded institutional-source pass

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.