Does artificial intelligence pose a threat to safe nursing care?

Turun yliopisto

In the end of November 2022 the world as we know it took I slight turn, as OpenAI released the prototype of ChatGPT, a chatbot and virtual assistant built upon language processing models trained with vast amounts of openly accessible data. The system made headlines internationally and provoked global debate of its potential usefulness, but also of the threat it may pose to the human science, culture and society. Four months later, an open letter to pause giant artificial intelligence (AI) experiments was released (1), and this May the World Health Organization issued a statement calling for safe and ethical AI for health (2). For the public and the policy makers, all this raises concerns regarding AI. Some may even wonder: if AI is implemented into nursing care, are we safe?

For many, AI is a synonym for intelligent systems such as ChatGPT able to pass the Turing test, or in more general terms, to pass as a human being. In nursing, the most notable advancements in AI to the public are the advancements in robotics, as today’s assistive robotic systems can for example provide information and guidance, transport materials and medication or aid communication (3). However, AI is not just robotics or ChatGPT, but an umbrella term for a myriad of systems designed for complex reasoning tasks. The main objective of AI research for clinical nursing is not to create systems to replace the nursing workforce, but to develop systems to support nurse decision-making, to assist in documentation or to monitor the patient, to name a few (4). At its best, AI systems developed for nursing would help nurses in providing the best possible care with minimal interruptions.

There is no question that we should tread lightly if we are to implement novel AI-based models like the highly advanced language models into nursing care. But how can we ensure that AI systems developed in nursing are, in fact, safe? The development of AI systems should be user-based and conducted within collaborative teams with representation of nurses, nurse informaticians and nursing researchers. Ethical concerns should be considered within all levels of the development process, including assessing the need for developing the system, the quality and possible bias presented by the datasets used to train the models, and the outcomes given by these systems. As with all novel technologies implemented into nursing care, the adoption of AI should be complemented with comprehensive nursing education on how the systems work, how to interpret the information given by them, and what the potential risks are when using them. Only through knowledge, the safe use of AI in clinical nursing care is possible.

Hanna von Gerich
The writer is a doctoral researcher at the Department of Nursing Science in University of Turku. Her research interests include combining secondary data sources with artificial intelligence-based technologies in the assessment of nursing care.

  • Future of life Institute. Pause Giant AI Experiments: An Open Letter. [online document]. Available from: https://futureoflife.org/open-letter/pause-giant-ai-experiments/. Published 22.3.2023. Accessed 19.5.2023
  • World Health Organization. WHO calls for safe and ethical AI for health. [online document]. Available from: https://www.who.int/news/item/16-05-2023-who-calls-for-safe-and-ethical-ai-for-health. Published 16.5.2023. Accessed 19.5.2023.
  • Ohneberg, C., Stöbich, N., Warmbein, A. et al. Assistive robotic systems in nursing care: a scoping review. BMC Nurs 22, 72 (2023). https://doi.org/10.1186/s12912-023-01230-y
  • von Gerich H, Moen H, Block LJ, Chu CH, DeForest H, Hobensack M, Michalowski M, Mitchell J, Nibber R, Olalia MA, Pruinelli L, Ronquillo CE, Topaz M, Peltonen LM. Artificial Intelligence -based technologies in nursing: A scoping literature review of the evidence.International Journal of Nursing Studies. 2022; 127. https://doi.org/10.1016/j.ijnurstu.2021.104153.