A recent survey demonstrates that a considerable number of general practitioners (GPs) in the UK are using generative AI tools, like ChatGPT, in their everyday clinical practices. This trend points to the swiftly increasing influence of artificial intelligence in the healthcare sector—an advancement that could transform how patients receive care, while also raising important ethical and safety issues.
A recent study conducted by researchers at Uppsala University in Sweden indicates that a considerable number of UK general practitioners (GPs) are incorporating generative AI tools, such as ChatGPT, into their clinical workflows. This highlights the swiftly increasing role of artificial intelligence in healthcare—a progression that holds the promise to transform patient care but also introduces significant ethical and safety challenges.
“Despite the ongoing discussion about the hype surrounding AI, our research indicates that the implementation of AI in healthcare is not a distant prospect—it’s occurring right now. Physicians are finding these tools beneficial. The medical community needs to take immediate action to address the ethical and practical challenges that generative AI presents for patients,” explains Dr. Charlotte Blease, the lead researcher and Associate Professor at Uppsala University.
The findings show that 20 percent of GPs reported the use of generative AI tools in their practices, with ChatGPT being the most commonly utilized. Collaborating with partners from Harvard Medical School in Boston, USA, and the University of Basel in Switzerland, this is the most thorough investigation of generative AI’s role in clinical settings since the introduction of ChatGPT in November 2022.
The research was carried out in February 2024 as part of a monthly omnibus survey and aimed to include GPs from various regions across the UK. A total of 1,006 GPs associated with Doctors.net.uk, the largest professional network for UK doctors, were surveyed.
The purpose of this study was to assess how widespread the use of AI-powered chatbots is among GPs in the UK and to explore how these tools are being applied in clinical environments. With the emergence of large language models (LLMs), there has been significant interest in their capacity to assist medical professionals in tasks such as documentation and differential diagnosis.
The study not only highlighted that 20 percent of GPs utilized AI tools but also revealed that among those using these technologies, 29 percent leveraged them for creating documentation after patient consultations, and 28 percent used them to aid in differential diagnosis.
These insights indicate that AI chatbots are becoming essential tools in healthcare, especially in alleviating administrative tasks and enhancing clinical decision-making. However, the use of such tools entails risks. The possibility of AI generating incorrect information (“hallucinations”), worsening existing biases, and threatening patient privacy is considerable. As these technologies develop, there is an urgent need for the healthcare sector to develop comprehensive guidelines and training programs to ensure their safe and effective deployment.
“This research highlights the increasing dependence of UK GPs on AI tools, despite the absence of formal training, guidance, and the associated risks. As the healthcare field and regulatory bodies continue to face these challenges, the demand for training physicians to thrive in the 21st century is more crucial than ever,” concludes Blease.
