Trustworthy AI performance through daily AI impact assessment platform
The prevalence of Artificial Intelligence (AI) and related technologies can be seen in business and society, as well as healthcare. The complexity and abundance of data in the healthcare sector offers space for AI to increasingly be used in the field by healthcare providers and carers. With the increased use of AI comes the increased importance of analyzing and assessing the governance aspects and societal, ethical, legal and health impact performances of AI systems in the health sector. YAGHMA’s AI impact assessment tool aims to assess these impacts to improve current and future AI towards trustworthy AI in healthcare by customizing performance goals, priorities, and implementation practices throughout the lifecycle of AI systems. The adoption of the impact assessment can facilitate future learning and development for health professionals and overall sector to better address AI’s ethical (and others) impacts on society.
AI IMPACT ASSESSMENT FOR HEALTHCARE
YAGHMA’s AI impact assessment tool undergoes five key steps. The first step includes conducting a materiality assessment (see Material Issues in the Blossoming World of AI for further explanation on material issues) to best understand which issues to prioritize and which issues are most important both for the healthcare sector and the company designing, developing, deploying or/and using the AI system. As each application of AI differs in priorities, it is necessary to customize the analysis and assessment to each company and project. In the second step and by using the outcomes of the materiality assessment, YAGHMA contextualizes the identified material issues to the companies AI system. These include the ethical, legal, social and health priorities specific to the AI. The third step conducts a dynamic AI impact assessment and update AI project with deeper insights into its priority impact performance, risks, and opportunities.
The results will hereafter allow YAGHMA to provide suggestions for improvement, integrating these into a tailor made AI project strategy. The fifth and final step is to audit/monitor the AI system on a periodically basis to ensure the AI system remains trustworthy and in line with health sector priorities.
AI Impact Assessment tool paves the way for healthcare professionals, policy makers and other healthcare stakeholders to make informed decisions on AI’s societal impacts. Take a look at the ‘AI Taxonomy’ – one of the results from our AI Impact Assessment Tool.
