Responsible AI

Governance by design is an approach where the governance model and its rules are built into the design system itself, from the initial experiment phase of AI initiatives through to full operational use. We view this model for governance as an enabler of value creation, providing security and assessment frameworks necessary for our business to move quickly and confidently. This model adapts the robust governance practices already in place to address the unique challenges and opportunities GenAI presents.

We apply our robust Risk Management Framework. From a privacy perspective, this means processes and standards such as our Global Privacy Statement, Client Data Privacy Principles and Privacy Impact Assessments, with enterprise oversight by our Chief Privacy and Data Ethics Officer are applied.

In each jurisdiction in which we operate, we comply with local laws, regulatory requirements, guidance, and standards applicable to responsible data use, including those relating to safety and security, fairness and non-discrimination, transparency and explainability, human oversight and control, accountability, beneficial use, and societal and environmental impact. Please also refer to Privacy and data security and Responsible data use and AI.

We recognize that using AI in our business may lead to environmental impacts. We are increasing our understanding of these impacts. In 2025, we started measuring the carbon footprint of our supply chain, reporting it as a new category of carbon emissions in our emissions inventory. This data collection milestone creates a pathway toward identifying the emissions associated with AI use.

As we implement GenAI across our business, we aspire for it to be a net positive for our workforce by supporting them, removing repetitive tasks and manual work, and opening pathways for employees to increasingly engage in work that brings added value.

Our internal employee upskilling programs emphasize the importance of maintaining human oversight, validating the accuracy of GenAI outputs and avoiding the use of sensitive information when prompting. As AI continues to evolve, we will continue to proactively refine our guidelines and safeguards as needed to uphold our commitment to operate as a trusted and responsible business.