Interview with Arundhati Bhattacharya,
President and CEO, Salesforce South Asia

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?

Having led the digital transformation of one of India’s most trusted financial institutions, I’ve seen firsthand that trust cannot be assumed—it must be earned continuously. For legacy institutions, adopting AI isn’t just a technological shift; it’s a cultural one. It requires transparent communication, strong internal governance, and a deep respect for customer data.

At SBI, when we moved to mobile banking and AI-enabled fraud detection systems, we did so by engaging every stakeholder—from board members to branch managers—on the ‘why’ behind the change. That same mindset applies today. Institutions must educate customers, create fallback mechanisms for AI decisions, and ensure that human oversight remains integral in high-stakes processes like lending, healthcare, or public services. In markets like Sri Lanka, a strong foundation of transparency and data stewardship is critical, as they forward in their digital transformation journey.

What governance frameworks or accountability mechanisms do you believe should be industry-standard when deploying AI at scale?

AI systems must be understandable—not just to data scientists, but to regulators, end users, and the general public. At Salesforce, we advocate for robust documentation and explainability of models so that any decision can be traced and justified.

Industry-wide, we need common accountability mechanisms such as bias detection audits, privacy-by-design protocols, and redressal frameworks for unintended outcomes. For example, AI used in recruitment must be tested for demographic fairness, and healthcare algorithms must be validated across diverse populations.

I’m encouraged to see countries like Sri Lanka incorporating responsible AI principles in their National AI Strategy. As businesses, we must complement these efforts with voluntary guardrails that often go beyond regulation.

In emerging economies like India, where digital literacy and access are uneven, how can organizations ensure that AI doesn’t widen the trust gap or exacerbate bias?

This is a deeply important question, because AI reflects the data and assumptions we feed into it. In growing economies where digital access and representation are uneven, bias isn’t just a risk—it’s inevitable unless addressed with clear intention and care.

Organizations must first ask: Who is represented in the data? Who is being left out? And then take action—by partnering with local communities, diversifying data sources, and ensuring interfaces are accessible in local languages. At Salesforce, we support initiatives that focus on inclusive design and multilingual AI, which are particularly important in South Asia’s linguistically rich environments.

Moreover, AI literacy must become part of our skilling agenda. Through initiatives like Trailhead and our Virtual Internship Program in India, we’re helping students and early professionals build both technical and ethical fluency in AI. Equity in access and equity in design must go hand-in-hand.

Trust is often earned through leadership. What ethical principles guide your own decisions when championing new technologies across Salesforce’s South Asia operations?

In today’s world, where technology is reshaping every facet of life and work, I believe leadership must be anchored in conscience as much as competence. For me, technology is only meaningful when it uplifts people—when it bridges gaps, not widens them. At Salesforce, I’ve always been guided by five enduring principles: empathy, transparency, inclusion, accountability, and long-term societal impact.

I ask myself simple but profound questions: Will this technology empower the underserved? Will it reinforce or reduce inequities? Will it be understandable, accessible, and safe for those who rely on it? These are not just philosophical reflections—they translate into the operational and strategic choices we make every day.

At Salesforce, I deeply resonate with our commitment to being Responsible by Design. This means trust isn’t an afterthought—it is architected into our systems, governance, and culture. From the Einstein Trust Layer to our AI Acceptable Use Policy, we embed ethical considerations at every level. And it goes beyond the technology stack—it extends to how we train our teams, support our customers, and hold ourselves accountable when things go wrong.

In the South Asia context, where institutions are at different stages of digital maturity, the role of leadership becomes even more critical. Trust is not built in boardrooms—it’s built in how responsibly we deploy innovation in hospitals, schools, government programs, and small businesses. As leaders, we must embody the values we want our technologies to reflect. Because in the end, the legacy of our leadership won’t be the code we write—it will be the trust we inspire.

(Arundhati Bhattacharya is the President and CEO of Salesforce South Asia. A Padma Shri awardee and the first woman to chair the State Bank of India, she brings over four decades of leadership across finance and technology. Recognized globally by Forbes and Fortune as one of the world’s most powerful business leaders, she now spearheads Salesforce’s strategic growth and ecosystem engagement across the region.)