designing responsible AI systems
Case Studies
The Brief
The Responsible AI (RAI) Index measures and tracks how well organisations are designing and implementing RAI systems, with a view to fairness, accountability, transparency, and impact on people and society.
The Responsible AI (RAI) Index measures and tracks how well organisations are designing and implementing RAI systems, with a view to fairness, accountability, transparency, and impact on people and society.
The approach
An online quantitative survey with 400+ business decision makers who have significant influence over AI strategy working in Australian organisations with at least 20 employees. All represented organisations that have deployed AI in their business, or were planning to do so in the next 12 months. A range of business sizes, industries and locations were included in the study to ensure an accurate representation of the Australian business landscape.
An online quantitative survey with 400+ business decision makers who have significant influence over AI strategy working in Australian organisations with at least 20 employees. All represented organisations that have deployed AI in their business, or were planning to do so in the next 12 months. A range of business sizes, industries and locations were included in the study to ensure an accurate representation of the Australian business landscape.
The challenge
The challenge was to design a robust measure that captured the various aspects of responsible AI, including strategy & leadership, governance, data and security, people and skills, and monitoring and review. The resulting Index incorporates self-assessed organisational performance scores and practices and actions that have been implemented on each aspect of RAI.
The challenge was to design a robust measure that captured the various aspects of responsible AI, including strategy & leadership, governance, data and security, people and skills, and monitoring and review. The resulting Index incorporates self-assessed organisational performance scores and practices and actions that have been implemented on each aspect of RAI.
The insight
The 2022 RAI Index exposes a worrying ‘action gap,’ with the results revealing a significant gap between awareness and action for most organisations. 82% of organisations are confident they are taking a best-practice approach to AI, but on closer inspection, only (24%) are actually taking deliberate actions to ensure their own AI systems are developed responsibly.
Positively, compared with 2021, more organisations are taking an enterprise-wide approach to the development of AI.
The 2022 RAI Index exposes a worrying ‘action gap,’ with the results revealing a significant gap between awareness and action for most organisations. 82% of organisations are confident they are taking a best-practice approach to AI, but on closer inspection, only (24%) are actually taking deliberate actions to ensure their own AI systems are developed responsibly.
Positively, compared with 2021, more organisations are taking an enterprise-wide approach to the development of AI.
The outcome
The 2022 Responsible AI Index and its findings garnered widespread media coverage, sparking conversation about the design and implementation of RAI. Specifically, it helps to support ethical and transparent AI practices, improve the quality of data driven decision making, enhance compliance, reduce bias and discrimination, and improve data privacy and security.
The 2022 Responsible AI Index and its findings garnered widespread media coverage, sparking conversation about the design and implementation of RAI. Specifically, it helps to support ethical and transparent AI practices, improve the quality of data driven decision making, enhance compliance, reduce bias and discrimination, and improve data privacy and security.
