Viewing all of the ethical and existential risks associated with Artificial Intelligence we should encourage AI researchers and developers to develop frameworks for implementing AI Governance in their respective organizations. AI Governance can learn much from the Data Governance programs in IT. Data Governance has been around since the 90’s and has mature processes and tools.
Each organization should establish an AI Governance Steering committee consisting of director level managers who set direction and review high level progress. The AI Governance Council would be responsible for making operational decisions regarding the AI product and for initiating projects that touch upon ethical issues.
Organizations should conduct an AI maturity assessment to determine the level of ethical AI they have achieved. Some of the AI Governance processes should recur, like validating model performance and reviewing emergent responses. The list of tools used to govern AI will be specific to each organization. IBM’s WatsonX.governance platform is one example.
