Ethics of Clinical AI

4-Lecture Video Series — Slide Decks

RevealJS lecture slides for the Ethics of Clinical AI series. Covers opacity and accountability, prediction and human oversight, fairness and monitoring, semantic infrastructure, and responsible AI governance for military trauma.
Modified

June 9, 2026

About This Series

These slide decks accompany the Ethics of Clinical AI blog series at Data InDeed.

This series is for trauma registry analysts, data scientists, clinical informaticists, and military medical professionals who need a rigorous, operationally grounded framework for evaluating the ethics of clinical AI — from registry design through model deployment and governance.


Lecture Schedule

# Title Posts Slides ~Duration
01 Opacity, Accountability & Ethical Failure Modes in Clinical AI 01 · 02 · 03 55 min
02 Prediction, Human Oversight & The Ethics of Data Exclusion 04 · 05 · 06 55 min
03 Fairness, Performance Monitoring & The Ethics of Automation 07 · 08 · 09 55 min
04 Semantic Infrastructure, Responsible AI & DoDTR Modernization 10 · 11 · 12 55 min

Series Overview

Lecture 1 — Opacity, Accountability & Ethical Failure Modes Challenges the reflexive demand for interpretability. When are black boxes ethically defensible? What does accountability actually require? And where do ethical failures in registry-based AI really originate — hint: it’s upstream of the model.

Lecture 2 — Prediction, Human Oversight & Data Exclusion Risk scores predict — they don’t decide, justify, or bear responsibility. Human-in-the-loop is not a safeguard when the workflow makes meaningful review impossible. “Messy” patients are disproportionately the most clinically important.

Lecture 3 — Fairness, Monitoring & Automation Fairness begins at data collection, not the loss function. Monitoring is an ethical requirement, not optional governance overhead. Automating CPG compliance monitoring faster is ethically valuable — but only with appropriate audit infrastructure.

Lecture 4 — Semantic Infrastructure & Responsible AI Local codes are an ethical failure of design. Ontology governance is a political act. The Delphi consensus on responsible AI leaves six critical gaps. DoDTR modernization is not a technology upgrade — it is an ethical obligation with real consequences for real patients.


Companion Resource

The Ethics Series Blog Posts provide the full argument behind each lecture, with detailed references and case examples from military trauma and clinical registry analytics.


Instructor Resource

The Master Speaker Notes provide lecture-by-lecture teaching guidance: key talking points, timing recommendations, discussion prompts, and common student questions — organized by slide section.