AI Solutions Should Have a Positive Impact on Humanity – Prof. Jerry John Kponyo

Prof. Jerry John Kponyo, PI and Scientific Director of the Responsible AI Lab (RAIL), in contributing to a panel discussion on Safeguarding Privacy in Africa’s IT Landscape at the 5th Privacy Symposium Africa organised on the 8th November 2023 in Mauritius, indicated that AI solutions must be designed and deployed to have a positive impact on humanity.

According to Prof Kponyo, ethical AI is responsible AI, and AI is responsible to the extent that it is human-centred.

“A key component of developing AI solutions is the datasets based on which models are trained. Data must be representative and without bias. The processes involved in acquiring and managing data must respect the rights of people. The development of AI solutions goes through the envisioning, data, modelling, and deployment stages, ensuring that AI solutions are delivered with integrity and equity, respecting individuals, and always being mindful of their social impact. We recognise that AI is the new chip on the block, and the lab applies AI in Agriculture, Health, Energy and for the differently abled,” he said.

Prof. Kponyo introduced the FACETS framework, developed by RAIL, as a quantitative approach to measuring responsible AI. “FACETS considers a pipeline that captures the whole process from envisioning to deployment in measuring the performance of AI solutions and considers the various stages of the AI innovation process,” he added.

He said ethical AI should consider Fairness, Accountability, Confidentiality, Ethics, Transparency, and Safety measures, which the FACETS framework stands for.