AI is increasingly recognised for its potential to transform medicine. AI, a branch of computer science that simulates human intelligence, is already capable of performing tasks once thought to require human expertise. In allergy and immunology, early applications have shown promise: predicting asthma from electronic health records, diagnosing allergy using epigenetic data, and identifying eczema through image analysis.
Recent advances, particularly in deep neural networks and Large Language Models (LLMs), have accelerated the pace of discovery. Generative AI systems, such as ChatGPT, exemplify this new era, where text and image generation based on user prompts is now commonplace. These technologies highlight both the speed of progress and the scope for future innovation.
Despite nearly 60,000 AI-related medical publications in 2024, translation into clinical practice within allergy and immunology remains limited. By mid-2024, only a little over a thousand AI-enabled medical devices had received FDA approval – and none were specific to allergy or immunology. This gap raises pressing questions about why the specialty lags behind others.
Several factors contribute to this discrepancy. The first wave of AI adoption largely focused on imaging-heavy disciplines such as radiology and pathology, where digital workflows are well-established. Allergy and immunology, in contrast, rely more heavily on clinical and laboratory data, areas where fewer AI-based, FDA-approved applications exist. This has likely slowed uptake and reduced AI literacy across the field.
Yet the potential is undeniable. AI tools could support diagnosis, management, prevention, patient self-care, and research. Proof-of-concept studies demonstrate clear opportunities, but the real challenge lies in moving from theoretical promise to meaningful clinical impact.
To address this, experts propose a six-point roadmap: focusing research on impactful challenges, setting clinically relevant benchmarks, ensuring rigorous validation, enabling operational implementation, fostering trustworthy adoption, and embedding lifecycle management. Such a framework shifts emphasis from experimental studies to tangible improvements in patient care.
By adopting this roadmap, allergy and immunology can catch up with other specialties and harness AI to deliver safer, more effective, and more personalised treatment. The future of the field may well depend on bridging this crucial gap between research and clinical reality.
Reference
van Breugel M et al. Artificial Intelligence in allergy and immunology: recent developments, implementation challenges, and the road towards clinical impact. J Allergy Clin Immunol. 2025;DOI:10.1016/j.jaci.2025.08.022.
