{
  "abstract": "Digital tools powered by artificial intelligence (AI), particularly generative AI in the form of large language models, are becoming ubiquitous in clinical practice, from assisting documentation to augmenting clinical reasoning. While AI tools may enable clinicians to work more effectively by upskilling (enhancing existing skills) or reskilling (acquiring new skills), they could also risk clinicians becoming overdependent on them and losing existing skills (deskilling). For clinicians-in-training or early career clinicians, formative opportunities for developing skills for tasks now performed by AI may be lost completely (never-skilling). Such deskilling could imperil clinician autonomy and patient safety in the event of AI errors and failures. Here, we explore the empirical evidence of AI-induced deskilling, propose 10 mitigation strategies and suggest educational reforms that ensure clinicians use AI in ways that maintain clinical skills.",
  "authors": [
    {
      "affiliations": [
        "Digital Health and Informatics, Metro South Health Service District, Brisbane, Queensland, Australia",
        "Centre for Health Services Research, The University of Queensland, Brisbane, Queensland, Australia",
        "Queensland Digital Health Centre, The University of Queensland, Brisbane, Queensland, Australia"
      ],
      "name": "Ian Scott"
    },
    {
      "affiliations": [
        "Queensland Digital Health Centre, The University of Queensland, Brisbane, Queensland, Australia"
      ],
      "name": "Anton van der Vegt"
    },
    {
      "affiliations": [
        "Bioethics Centre, Monash University, Melbourne, Victoria, Australia"
      ],
      "name": "Peter Douglas"
    }
  ],
  "title": "How can we prevent clinical deskilling when using AI?",
  "uid": "0972748a-03b9-5414-a717-68b89b1be1d9"
}