Ethical AI and responsible leadership in emerging markets

From your vantage point leading Salesforce South Asia, how do you define “ethical AI” in the context of today’s rapidly digitizing businesses?

- 

Ethical AI, in my view, is not just about meeting compliance checklists—it’s about upholding values of trust, transparency, and inclusion throughout the lifecycle of AI design, development, and deployment. In South Asia, where digital transformation is gaining pace but foundational trust in emerging tech is still being built, ethical AI means being deliberate: building systems that are explainable, fair, and aligned to human-centered values.

At Salesforce, our approach to responsible AI is rooted in our core value of Trust. We've developed a robust governance framework, grounded in five key principles: responsibility, accountability, transparency, empowerment, and inclusiveness. These guide how we build and deploy AI technologies—ensuring they are safe, effective, and beneficial to all stakeholders.

Take our Einstein Trust Layer as an example—it ensures that sensitive customer data is never used to train generative AI models, and provides enterprise-grade safeguards like data masking, audit trails, and user-controlled prompts. It’s a practical embodiment of our ethical AI commitment—giving customers the confidence that innovation will not come at the cost of integrity.

Moreover, we believe that ethical AI isn’t just a product decision—it’s a leadership responsibility. That’s why we’ve established dedicated advisory councils, cross-functional governance bodies, and even bias evaluation protocols to continuously assess how our AI systems impact different groups. In an era where businesses are expected to do more than just grow revenue, ethical AI is not optional—it’s foundational to long-term, sustainable digital success.

Given your leadership at SBI during its digital pivot, how do legacy institutions build trust while adopting disruptive technologies like AI—especially when dealing with sensitive consumer data?