Throughout the history of mankind, medical discoveries such as antibiotics, vaccines, and anesthesia have transformed the healthcare industry. Today, the introduction and adoption of artificial intelligence (AI) are driving a digital transformation in healthcare. This evolution can potentially improve efficiency, reduce medical costs, and enhance patient care while creating high-quality solutions that promote healthier and longer lives. Since its inception, AI has made remarkable progress in healthcare, bringing about significant changes and numerous opportunities.
One notable advancement is the development of Manus AI, an autonomous AI agent that represents a significant leap forward in AI technology. This AI-powered solution redefines how medical professionals process healthcare data, diagnose diseases, and develop treatments. According to a survey report published by Nvidia in 2025, 63% of respondents reported actively using AI-powered tools in healthcare. Figure 1 illustrates the top three AI use cases currently prevalent in the healthcare industry.
AI has many important applications in the healthcare industry. Currently, it is being used for drug discovery and design, automated document processing, medical image recognition, and inventory tracking. For example, Exscientia, a drug design and development company, utilizes generative AI built on Amazon Web Services (AWS) to generate potential drug compounds in silico. Automated robots assist in developing these drug candidates. This AI-driven approach has enabled Exscientia to accelerate the drug development process by 70% and reduce capital costs by up to 80%. Notably, six molecules designed using this method have advanced to clinical trials.
In terms of automated medical documentation, DeepScribe’s AI-powered scribe system automatically creates personalized documentation for clinicians by capturing patient visits in real time. The system generates billable documentation and syncs it directly with the clinician’s electronic health record (EHR) system. Importantly, this system has reduced clinician burnout, improved revenue, and enhanced the quality of patient care.
The adoption of AI technology has also improved medical image recognition. A company called PathAI uses AI for this purpose. Traditionally, it can be challenging for expert pathologists to accurately interpret and diagnose pathology slides, even with specialized training and collaboration. Inaccurate interpretations can negatively impact diagnoses and drug development trials. PathAI employs a machine learning technique known as graph neural networks (GNNs) to reveal spatial relationships in tissue samples that may not be easily detected by visual inspection.
AI technology can also optimize inventory management systems in healthcare. For instance, Hospices Cibils de Lyon (HCDL) faced challenges in managing and tracking their implanted medical devices. IDENTI medical company provided a solution by developing TotalSense RFID cabinets, a mobile app, and handheld RFID scanners. These tools were integrated into the IDENTIPlatform, an AI-driven cloud-based software. As a result, the pharmacy department at HCDL can now effectively track all inventories once distributed across various hospital departments, maintaining visibility and control over their medical supplies and instruments.
Despite these advancements, several challenges hinder the widespread adoption of AI in healthcare. One of the major issues is the complexity of AI models, particularly deep learning models, which often demonstrate low interpretability. This “black box” nature obscures the internal workings of AI, making it difficult to understand how decisions are reached. A lack of transparency can lead to serious patient safety issues, such as misdiagnoses and inappropriate treatments.
Moreover, AI models in healthcare are typically trained on datasets comprising historical health records, treatment outcomes, and patient demographics. These datasets may contain biases due to the underrepresentation of certain groups or genders. For example, an AI algorithm developed by Optum and used in several U.S. health systems was found to display biased outcomes, prioritizing healthier white patients over sicker black patients. As a result, high-risk black patients were less likely to receive additional interventions despite having greater medical needs.
In summary, AI has the potential to significantly transform the healthcare industry, improving patient care, diagnostic accuracy, and cost management. However, challenges related to low interpretability and biases in AI models must be addressed. As AI becomes increasingly integrated into healthcare, it is essential to adopt this technology responsibly and uphold high ethical standards.