Augmented intelligence and dermatology—Part II: Bias, benchmarks, guidelines, ethics, regulation, and future directions - 18/12/25

Doi : 10.1016/j.jaad.2024.10.132

Edwin Dovigi, MD, MS ^a, Shannon Wongvibulsin, MD, PhD ^b, Ivy Lee, MD ^c,^⁎ , Rob Novoa, MD ^d, Zhuo Cai, MD ^d, Veronica Rotemberg, MD, PhD ^e, Roxana Daneshjou, MD, PhD ^d

^a Department of Dermatology, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania

^b Division of Dermatology, David Geffen School of Medicine, University of California, Los Angeles, Los Angeles, California

^c Comprehensive Dermatology Center, Pasadena, California

^d Department of Dermatology, Stanford University, Stanford, California

^e Dermatology Service, Memorial Sloan Kettering Cancer Center, New York, New York

^∗ Correspondence to: Ivy Lee, MD, Comprehensive Dermatology Center, 625 South Fair Oaks, Suite 200, Pasadena, CA 91105. Comprehensive Dermatology Center 625 South Fair Oaks Suite 200 Pasadena CA 91105

Abstract

Despite the proliferation of artificial intelligence (AI) tools, numerous challenges remain in the development and evaluation of AI-related technologies. Furthermore, a practice gap exists in how to responsibly implement AI effectively, ethically, and legally in dermatology. Part 2 of this CME review series will explain the key technical challenges and principles of trustworthy and responsible AI as well as describe the ethical challenges and review resources to help guide safe, effective, and ethical AI in dermatology. This content will empower learners to critically analyze these tools and determine whether an AI solution is appropriate for their specific clinical need, patient, and practice setting.

Le texte complet de cet article est disponible en PDF.

Key words : artificial intelligence, augmented intelligence, dermatology, ethics, digital health

Abbreviation used : AI

Plan

Introduction

Challenges in developing and evaluating AI-related technologies in dermatology

Data quality and standardization challenges

Clinical integration challenges

Evaluation and validation challenges

Core concepts about transparency and trustworthiness in AI

AI transparency

AI trustworthiness

Patient and clinician perspectives

Technical, regulatory, and ethical frameworks for AI in dermatology

Technical frameworks

Regulatory frameworks

Ethical frameworks

Reporting and implementation guidelines

Conclusion

Declaration of AI-assisted technologies in the writing process

	Drs Wongvibulsin and Dovigi are cofirst authors.
	Funding sources: None.
	Patient consent: Not applicable.
	IRB approval status: Not applicable.
	Prior publication: The contents of this manuscript have not been previously published and are not currently submitted elsewhere.
	Date of release: January 2026.
	Expiration date: January 2029.