Cindy (Xin) Tu, Director of IT and Data Audit at Discover Financial Services, speaks with Mudit Gupta, Partner, FS Consulting AI Practice Leader at EY, in a video interview about bridging the talent gap on AI and data, training and recruitment, the evolution of GenAI from theory to tool, governance, and learning through deployment.

Bridging the talent gap in AI and data

At the onset, Tu shares her perspective on the immense challenges organizations face in acquiring the right talent. As she puts it, “There’s a war on talent, especially when it comes to AI and data.”

Delving further, Tu emphasizes that professionals with expertise in areas like data audit and AI risk management are extremely hard to find. “I don’t think it exists yet,” she remarks, speaking about professionals with hands-on AI experience. “No one knows how to automate AI yet. I think we’re just trying to figure it out.”

A two-pronged Strategy — Train and recruit

To address the talent shortfall, Tu advocates for a dual strategy focused on both training and recruiting. She notes that it is critical to talk about digital literacy and recruit talent from the marketplace.

This balanced approach recognizes that while hiring externally is necessary, investing in internal capability-building is just as crucial. Given the lack of ready-made experts, Tu emphasizes identifying individuals with a foundational understanding of data and AI — more importantly, those who are eager to learn. She says, “They need to be hungry.”

Additionally, Tu says, “You have to hire someone who is willing to learn and is a fast learner, who is motivated enough that… if you make the investment and try to teach them how to do this… they’re willing to invest the time to learn.”

Reiterating the importance of aligning recruitment strategies with internal development plans, Tu urges having a recruitment strategy. She further suggests identifying where the gap is and having a robust program in place to invest the time to get there.

GenAI goes mainstream — Productivity, risk, and responsible adoption

Moving forward, Tu shares that the adoption of GenAI has moved well beyond experimentation. “I think the emergent channel definitely sees that because of more and more availability of the GenAI use cases, and we’re not stuck at the theory level anymore.”

With real-world applications now demonstrating measurable productivity gains, particularly in programming and automation, businesses are rethinking how they structure their workforce, she adds.

Furthermore, Tu highlights how the integration of GenAI is no longer confined to experimental labs or tech enthusiasts. It is embedded in daily routines, she says.  

Tu affirms, “I don’t think this is hype anymore. It’s becoming the mainstream. Everyone is using it. Everyone is talking about it.”

Governance is catching up

With increasing adoption, the focus has naturally shifted toward responsible use. Tu states that governance frameworks around AI, particularly in regulated industries like financial services, are maturing.

She recommends organizations categorize GenAI use cases in a way that allows for more targeted governance, she adds. This could involve adapting existing frameworks like IT governance, data governance, or model risk management to address new AI-driven scenarios.

Learning through AI deployment

Thereafter, Tu points out a critical evolution in how organizations are learning to manage AI risks — not just during testing, but after deployment. This iterative learning process, while sometimes reactive, is contributing to a more nuanced understanding of the risks involved and how to mitigate them.

Acknowledging the ongoing challenge of balancing innovation with responsibility, Tu says, “We’re still trying to balance that innovation versus risk management. So that’s a constant struggle.”

Still, she remains optimistic: “I do see more maturity compared with the use cases I’ve seen last year. So I think we’re on the right path.”

CDO Magazine appreciates Cindy (Xin) Tu for sharing her insights with our global community.