Dr. Arif Hussain, Chairperson of the School of Life Sciences at Manipal Academy of Higher Education (MAHE), Dubai
In the UAE alone, cancer incidence is projected to rise by 36% over the next decade, underscoring the urgent need for innovation in diagnosis and care. The Lancet Oncology recently highlighted a 60% surge in new cancer cases in the UAE between 2019 and 2023, reflecting broader regional and global trends. As healthcare systems grapple with rising patient volumes and complexity, artificial intelligence (AI) emerges as a potential game-changer in oncology. Around the world, healthcare systems are increasingly turning to artificial intelligence to reduce inefficiencies while managing growing patient demands. From administrative automation to advanced clinical support, AI is being woven into the fabric of modern healthcare, and this shift is not just conceptual. Market forecasts reflect this too. A report by GlobalData projects that global revenue for AI platforms in healthcare will reach nearly $19 billion by 2027, indicating a swift rise in the sector’s promise to adopt advanced technologies.
Among medical specialties, oncology stands out as one where the need for AI is both the most urgent and most intricate, being a field that generates immense volumes of data, from medical imaging and pathology to genomics and clinical notes. Clinical research shows that AI systems can match or even surpass expert performance in specific diagnostic tasks, including detecting malignant lesions in mammograms or identifying abnormal cells in pathology slides.
The Real World of Health Care
Despite this need, AI has yet to become a routine part of cancer care as there remains a significant lag between innovation in the lab and adoption in real-world clinical settings.
Most oncologists continue to rely on established diagnostic and treatment protocols. In practice, AI tools are still viewed as supplemental aids rather than core components of cancer care, therefore the technology remains largely on the periphery of clinical workflows.
Barriers to AI Integration in Oncology – A Multifaceted Challenge
While the potential of AI in oncology is widely recognized, its impact remains largely confined to pilot programs and research settings. Challenges like inconsistent data quality, poor system integration, and persistent concerns around clinical validation and trust continue to limit broader adoption. But what exactly are these barriers, and why are they so difficult to overcome?
A foundational barrier to effective AI implementation lies in the availability and quality of data. The success of robust AI models depends on large volumes of clean, well-annotated datasets, yet the complexity of cancer care makes this exceptionally difficult to achieve. Given the broad variability across cancer types, stages of disease progression, and patient populations, there is a substantial degree of diversity within the data inputs…
Beyond data limitations, regulatory and validation framework hurdles also slow AI adoption in oncology. Clinical tools must undergo rigorous testing to prove safety and effectiveness, but this process is often slow, expensive, and poorly suited to AI systems that continuously learn and adapt, casting doubt on whether the investment is justified. At the same time, economic and structural misalignments present further challenges. High upfront costs, unclear reimbursement, and outdated billing models, make it difficult for organizations to invest in AI, even when long-term benefits are clear.
Regulatory bodies such as the FDA are actively working to adapt their guidelines for these dynamic tools, but questions remain around how to ensure ongoing performance monitoring- especially post-deployment. Even after regulatory clearance, AI systems require continuous validation across diverse patient populations to safeguard against drift and maintain clinical relevance.
Integration into the clinical process
Integrating AI into real-world oncology practice introduces further complexity. Many healthcare institutions lack the necessary digital infrastructure, including high-performance computing capabilities, network bandwidth, data storage, and trained technical personnel to deploy and maintain AI systems effectively. Even when tools are available, if they are not seamlessly embedded into clinical workflows, they risk becoming burdensome rather than beneficial, contributing to clinician fatigue – an issue already prevalent in the region. According to a 2025 Strategy& (PwC Middle East) report, nearly 50% of healthcare professionals worldwide, including those in the GCC, are affected by burnout, largely driven by long hours and high-intensity work environments. Against this backdrop, poorly implemented AI risks adding to time pressure rather than alleviating it.
Maximizing the benefits of AI without exacerbating existing challenges will require equipping clinicians with the skills to interpret and use these tools effectively. For this, there is a need for medical education to adapt by integrating AI literacy and practical application as essential components of training.
Considering all these factors, trust emerges as the key challenge, with many AI systems operating as opaque “black boxes” that provide recommendations without transparent explanations. This lack of clarity can lead to hesitation among clinicians, particularly when decisions carry significant risk, but can also leave patients feeling uncertain about treatments driven by technology they don’t fully understand. Additionally, unresolved ethical and legal questions around accountability in cases of error deepen these concerns.
To address this, it is crucial that oncologists communicate AI-driven insights clearly and compassionately, especially when those insights impact life-changing decisions.
The Way Forward
Unlocking AI’s full potential in oncology requires coordinated collaboration among clinicians, researchers, industry innovators, regulators, and patient advocates to establish clear goals and best practices. Clinical trials must be thoughtfully designed to thoroughly evaluate AI tools across diverse real-world settings, focusing on outcomes that truly matter to both patients and providers. At the same time, medical education must evolve to equip future oncologists with the technical expertise and critical skills needed to effectively leverage AI.
One way this shift is already taking shape is at MAHE Dubai, where MSc Molecular Biology & Human Genetics, MSc Medical Biotechnology, and PhD students are exploring the integration of AI in oncology research – leveraging bioinformatics as a powerful tool for research and innovation.
Equally important are adaptive regulatory frameworks and innovative reimbursement models that encourage responsible AI innovation, while ensuring patient safety, data privacy, and equitable access.
While progress may be gradual, AI’s integration is poised to complement, not replace, the expert judgment and compassionate care provided by oncologists. Its greatest value lies in becoming a trusted ally through thoughtful, inclusive, and well-regulated adoption, shaped by the needs of patients and healthcare professionals alike. AI’s role in oncology will not be a sudden revolution, but a steady evolution, one that, with thoughtful adoption, will ultimately transform patient care for the better.
