>>> [!note] Migrated issue <!-- Drupal.org comment --> <!-- Migrated from issue #3606739. --> Reported by: [joshua1234511](https://www.drupal.org/user/3362218) Related to !6 >>> <!-- Drupal.org issue summary for: Add AI Empathy Healthcare Governance submodule Project: drupal/ai_empathy | Type: Feature request | Component: Code Paste the body below into the Drupal.org issue summary (Full HTML). --><h3 id="summary-problem-motivation">Problem/Motivation</h3> <p>AI Empathy Evaluation scores responses on empathy, decision accuracy and explanation quality. In healthcare &mdash; one of the highest-stakes domains for AI-assisted decisions &mdash; two governance dimensions matter just as much but are not currently measured:</p> <ul> <li><strong>Data protection</strong> of the sensitive patient data a clinical AI consumes (e.g. India's DPDP Act 2023, HIPAA, GDPR).</li> <li><strong>Clinical trust</strong>: whether a response is safe, transparent about uncertainty, and appropriately defers to clinician judgement rather than over-claiming.</li> </ul> <p>Popular surveys of real-world clinical AI (triage assistants, mental-health chatbots, readmission-risk prediction) routinely evaluate accuracy and speed while leaving these human-centred governance dimensions implicit. This makes the empathy instrument incomplete for clinical decision-support research and for organisations deploying AI in regulated medical settings.</p> <h4 id="summary-research-context">Research context</h4> <p>This submodule is the reference implementation backing an academic paper, <em>"Empathy, Trust, and Data Protection in AI-Assisted Clinical Decision-Making: A Governance Framework for Clinician Acceptance,"</em> being submitted to <strong>INDAM 2027</strong>, the conference of the Indian Academy of Management hosted by IIM Sambalpur (<a href="https://iimsambalpur.ac.in/indam/index.php">https://iimsambalpur.ac.in/indam/index.php</a>), under the <strong>Sustainability, Ethics &amp; Governance</strong> track. The module operationalises the paper's Empathy&ndash;Trust&ndash;Accountability (ETA) framework for medicine and produces per-response scores suitable for empirical analysis. It builds on prior peer-reviewed work, <em>"Evaluating Empathetic Decision-Making in AI: A Comparative Study of Open-Source Models in High-Stakes Scenarios"</em> (IJFMR, 2025, DOI <a href="https://doi.org/10.36948/ijfmr.2025.v07i06.63345">10.36948/ijfmr.2025.v07i06.63345</a>).</p> <h4 id="summary-steps-reproduce">Steps to reproduce</h4> <ol> <li>Install and configure AI Empathy Evaluation with a provider/model.</li> <li>Run an evaluation against any <em>medical</em> scenario.</li> <li>Observe that the result captures empathy/accuracy/explanation only &mdash; there is no signal for patient-data protection or for whether the response is safe and well-calibrated for clinical reliance.</li> </ol> <h3 id="summary-proposed-resolution">Proposed resolution</h3> <p>Add an optional submodule, <strong>AI Empathy Healthcare Governance</strong> (<code>ai_empathy_healthcare</code>), following the existing <code>ai_empathy_ccc</code> pattern (no hard coupling to the base module). It hooks the base module's evaluation events and adds two metrics for in-scope scenarios:</p> <ul> <li><strong>Data Protection (1&ndash;5):</strong> compliance of the response with a configurable framework &mdash; <strong>DPDP 2023 / HIPAA / GDPR</strong> &mdash; covering lawful purpose, consent, data minimisation, confidentiality and patient rights.</li> <li><strong>Clinical Trust (1&ndash;5):</strong> safety, transparency and calibrated confidence (communicating uncertainty, deferring to clinical judgement).</li> </ul> <p>Mechanism:</p> <ul> <li><code>PROMPT_ALTER</code> &rarr; frames the scenario as AI-assisted clinical decision support that must respect the configured data-protection framework.</li> <li><code>SCORE_ALTER</code> &rarr; a <code>HealthcareComplianceScorer</code> returns both metrics in a single scoring call; the base evaluator persists any returned score that maps to a result field.</li> </ul> <p>Scoring reuses the provider/model already configured in <code>ai_empathy.settings</code>, so no extra AI configuration is required. Scope is configurable (by scenario category, default <em>medical</em>, or all scenarios), and out-of-scope scenarios are never altered. The base <em>decision accuracy</em> metric acts as a prediction control, giving an Empathy &times; Data-Protection &times; Trust instrument at constant predictive quality.</p> <h3 id="summary-remaining-tasks">Remaining tasks</h3> <ul> <li>Review submodule code and naming.</li> <li>Confirm DPDP/HIPAA/GDPR guidance text is accurate and appropriately scoped (illustrative, not legal advice).</li> <li>Add a Kernel/Functional test covering the install (base-field creation) and the gating behaviour, in addition to the existing unit tests.</li> <li>Decide whether the two metrics should surface on the results dashboard and in any CSV/Views export.</li> <li>Documentation page for the submodule.</li> </ul> <h3 id="summary-ui-changes">User interface changes</h3> <ul> <li>New settings form at <code>/admin/config/ai/empathy/healthcare</code> (permission: <em>administer ai empathy</em>) to set the data-protection framework, scope, and reporting thresholds.</li> <li>New menu link "Healthcare governance" under AI Empathy Evaluation.</li> <li>Evaluation results for in-scope scenarios additionally carry Data Protection and Clinical Trust scores.</li> <li>Ships three clinical-AI example scenarios: AI Virtual Triage Assistant, Mental-Health Support Chatbot Disclosure, Readmission-Risk Prediction Disclosure.</li> </ul> <h3 id="summary-api-changes">API changes</h3> <p>None to the base module. The submodule consumes existing public extension points only (<code>AiEmpathyEvents::PROMPT_ALTER</code> and <code>AiEmpathyEvents::SCORE_ALTER</code>). New public classes are added under the <code>Drupal\ai_empathy_healthcare</code> namespace (<code>HealthcareComplianceScorer</code> service, <code>RegulationFramework</code> enum, event subscriber).</p> <h3 id="summary-data-model-changes">Data model changes</h3> <p>Adds two <code>float</code> base fields to the existing <code>ai_empathy_result</code> entity via <code>hook_entity_base_field_info()</code>: <code>data_protection</code> and <code>clinical_trust</code> (default 0). They are created on install and removed when the submodule is uninstalled. Adds one config object, <code>ai_empathy_healthcare.settings</code> (with schema). No changes to base-module config entities.</p>