Challenges and Barriers in Implementing AI for Clinical Applications in Anatomic Pathology

The transformation of anatomic pathology from microscope-based practice to digital and computational workflows has created unprecedented opportunities for artificial intelligence (AI)-driven diagnostics. Whole slide imaging systems, digital cytology, and enterprise pathology platforms now enable large-scale image analysis, quantitative tissue characterization, and integration with molecular and clinical data. In parallel, advances in machine and deep learning, and generative models have produced AI systems capable of tumor detection, grading, prognostication, biomarker assessment, and molecular inference directly from routine histologic and cytologic slides. Despite this rapid progress, translation of AI from research environments into routine clinical anatomic pathology remains limited. Barriers include data heterogeneity and limited generalizability, challenges in validation and regulatory compliance, infrastructure and interoperability constraints, workflow integration difficulties, and concerns regarding transparency, accountability, and professional trust. This review synthesizes current evidence on digital pathology and AI applications across surgical pathology and cytopathology, examines the technical, organizational, and regulatory factors that impede clinical adoption, and outlines practical recommendations for developing clinically deployable AI systems. Addressing these challenges through robust digital infrastructure, representative data sets, rigorous validation, and coordinated governance will be essential for realizing the full clinical potential of AI in anatomic pathology.