WA health entities must ensure that AI systems enable fair and equitable access to health care and government services and do not exacerbate existing disparities in access such as those related to race, sex, gender or socioeconomic status. Staff working with AI are adequately trained or qualified to monitor and manage bias in the algorithms, including self-developed models, which must be impartial toward race, sex, gender, religion, culture, language, ethnicity or any other ground; in the data used to train models, which must suit the WA health context and use case including its ethical, societal, cultural and legal context; and in the outputs, which are continuously monitored for compliance with anti-discrimination law and for disproportionate impact on demographic groups. Staff implementing AI are mindful of the harms of racial and gender bias and put mitigation strategies in place, consistent with MP 0185/24 Health Equity Impact Statement and Declaration Policy, MP 0160/21 Aboriginal Health Impact Statement and Declaration Policy and the NSQHS Standards user guide for Aboriginal and Torres Strait Islander health. Where AI listens to staff, patients or customers, comprehensive testing across the range of accents and languages is demonstrated. Transparency is addressed under Principle 6 where language barriers exist, and staff are aware of the National Agreement on Closing the Gap and its expectations for the transparency, accountability, access, sharing and control of Aboriginal data.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.