Queensland Health Artificial Intelligence Framework P1: Guiding principles: prioritise and design AI to serve the quintuple aim, improving patient outcomes through safe, effective solutions that reduce human bias, using resources efficiently with demonstrated value for money and no duplication, supporting workforce wellbeing by automating repetitive and augmenting complex tasks and personalising training, improving consumer experience through digital tools, and driving equity by designing for marginalised and regional and remote communities from the start and validating algorithms locally; and apply Australia's eight AI ethics principles (human, societal and environmental wellbeing, human-centred values, fairness, privacy protection and security, reliability and safety, transparency and explainability, contestability, accountability) in developing and using AI

What does P1 require?

The guiding principles, developed with internal and external stakeholders and endorsed in March 2024, are the foundational considerations for adopting AI in Queensland Health. In line with the Clinical Senate's June 2023 recommendations, Queensland Health prioritises AI projects that directly support the quintuple aim of health care. AI solutions must be safe and effective, reduce the impact of human bias and improve patient outcomes; show that their implementation is effective, value-adding, non-duplicative and value for money; support workforce wellbeing, satisfaction and retention by automating repetitive tasks or augmenting complex ones and by personalising education and training; empower consumers to use digital health tools to manage their health; and drive equity by being designed with people from marginalised communities and areas in mind from the start rather than retrofitted, with applicability in regional and remote communities ensured by validating algorithms in the local setting. Staff developing or using AI must also apply Australia's eight AI ethics principles published by the Department of Industry, Science and Resources and commit to ethical AI practice: human, societal and environmental wellbeing; human-centred values respecting rights, diversity and autonomy; fairness without unfair discrimination; privacy protection and data security; reliability and safety in accordance with intended purpose; transparency and responsible disclosure so people know when AI significantly affects or engages them; contestability through a timely challenge process; and accountability, with identifiable responsible people and human oversight. Alignment with the principles is assessed through the AI Risk Assessment Tool at each stage.