Do Androids Dream of Lived Experience? Navigating Risks, Uncertainty, and Best Practices Around Generative AI in...

In this viewpoint paper, we, as professionals working in patient and public involvement and engagement (PPIE), argue that the use of generative AI poses risks to the place of lived experience in PPIE because of its ability to generate outputs that can potentially eclipse those of people with lived experience. Generative AI appears to offer solutions to the National Health Service (NHS) and health research in an era of change and uncertainty-overstretched health infrastructure, unclear leadership, burned-out staff, and cuts and redundancies. NHS leaders are also pursuing the integration of generative AI into the NHS, in part to realize its "analogue to digital" plans. But those solutions come with commensurate risks. In this context, we have authored this viewpoint, aiming to (1) explain the risks that generative AI poses to those working with lived experience in PPIE contexts (distorted and biased outputs, contributing to inequalities, minimizing or replacing patient voices, and limiting learning); (2) clarify the positive opportunities that generative AI offers to PPIE (speeding up diagnostics, accessibility accommodations); and (3) provide a checklist that gives professionals clear pragmatic advice on using generative AI ethically in PPIE.