TD rolls out enterprise-wide responsible AI principles to bolster transparency, fairness

TD introduced enterprise-wide Responsible AI Principles to guide AI development, deployment and use across the bank.

The framework targets fairness, transparency, explainability, accountability, privacy, reliability and security across the AI lifecycle.

AI use cases must pass explainability, fairness and performance assessments before deployment, with ongoing monitoring once live.

Governance includes model risk assessments, privacy impact assessments and independent oversight reviews under an AI Risk Management framework.

A Trustworthy AI team operationalizes the policy, tracking performance, fairness and explainability to keep systems aligned with stated standards.