Xin Tu on Leading AI and Data Governance

In this edition of The Executive Outlook, we had the privilege of speaking with Xin Tu. With over 18 years of experience in IT, risk management, and auditing, including a decade in financial services, Xin has seen how data and AI can transform organizations, drive smarter decisions, and create resilient systems. 

Xin’s journey is not about chasing the latest technology trends. It’s about building trust, innovating responsibly, and ensuring AI is used ethically. Her story demonstrates that curiosity, discipline, and a commitment to governance, ethics, and purpose are as important as technical skills.

From Risk Management to Data Leadership

Xin began her career long before AI became a mainstream topic. “I started in IT and risk management when AI was more theory than practice,” she recalls. “Even back then, I saw that technology wasn’t just changing systems; it was transforming how businesses operate, how decisions are made, and how risk is evaluated.” 

Over time, Xin realized a critical truth: AI is only as good as the data it relies on. “If the data feeding a model isn’t accurate,” she explains, “the outcomes will never be reliable. That became my guiding principle in how I approach data and AI.” 

This understanding drew her into data governance, the foundation for trustworthy AI. “Governance is not a checkbox,” she says. “It determines how confidently we can use data to make decisions. It’s what allows innovation to happen safely.” 

Xin is also deeply involved in industry-wide governance efforts. She studies frameworks like DAMA, DCAM, and especially CDMC—the only cloud data governance framework by the EDM Council. She actively contributes to CDMC working groups, where AI governance principles are now being incorporated into the framework.

Prefer to listen on the go? Tune in to the full podcast episode on Spotify below:

A Decade of Change in Financial Services

Looking back on her ten years in financial services, Xin Tu smiles at all the changes she has seen. “There’s never a dull moment,” she says. “Risk isn’t fixed, it changes all the time, and it keeps us learning something new every day.” 

She recalls how things used to be: auditors going through piles of paper, checking only small samples, and hoping nothing was missed. “It was slow, tiring, and mistakes could happen,” she remembers. 

Today, she says, technology makes a big difference. Automation and data tools let her teams look at all the data, spot patterns quickly, and make smarter choices. “We’re not just saving time,” she adds. “We’re discovering insights that help the organization handle risk better and become stronger.” 

She explains that modern audits no longer rely on sampling. Instead, her teams use automation, data analytics, and AI-driven anomaly detection to perform 100% of population testing—identifying risks faster and with greater accuracy than ever before.

Watch the full conversation on YouTube by clicking the link below:

When AI Became Real

While machine learning had been around for years, Xin felt the world shift in 2022. “Then came ChatGPT,” she said. “Suddenly, AI was no longer abstract, it was real. People could see how it could reshape daily life, work, and decisions. The pace of change became undeniable.” 

She paints a vivid picture of AI in action: doctors catching cancer earlier, delivery companies mapping the fastest routes, restaurants predicting what customers would want tomorrow. “AI isn’t distant, it’s part of every decision we make,” Xin notes. 

Her advice is both simple and inspiring: “Stay curious. Look for ways AI can make your work smarter, not just faster. Curiosity is your compass in a world that never stops evolving.”

Designing AI Governance That Works

When asked how organizations should set up governance, Xin shares a clear step-by-step approach:

1. Define Acceptable Risk: “Governance is about balancing innovation and control. You need to know what level of risk your organization is willing to accept.”
2. Identify Stakeholders: “Who owns the framework? Who approves AI projects? Who ensures compliance? Without clear ownership, governance fails.”
3. Map Every AI Use Case: “Every model, tool, or application should be mapped from development to deployment. Internally developed models are not the same as third-party solutions. Risk reviews need to be tailored accordingly.”

Xin distinguishes between three major classifications:

• GenAI vs. non-GenAI
• Internally developed models vs. third-party AI tools
• Models requiring model-risk review vs. those needing third-party, cybersecurity, or privacy assessments

Each path triggers different approval committees and different layers of IT, data, model, and third-party risk management.

4. Make Governance Precise, Not Generic: “A framework should be practical and actionable, not abstract. Precision ensures trust and usability.”

Xin emphasizes that governance is not just about rules, it’s about people. Without champions across the organization, frameworks remain documents rather than living practices.

The Power of Support and Incentives

Xin believes that governance fails without support and motivation. “You can design the best framework,” she says, “but if people don’t use it, it fails.” 

She advocates for creating “data champions”, individuals who promote best practices, inspire their peers, and build a culture of trust. “Support matters, but incentives are equally critical. Tie governance goals to performance reviews. Reward those who embrace best practices and hold accountable those who resist. That’s how frameworks become effective in reality.”

She adds that strong governance requires clear KPIs and KRIs to measure whether controls are actually implemented, especially across different data classes (sensitive vs. confidential vs. general). Metrics also reveal which business units support governance and which ones resist it. 

Xin also stresses the importance of talent. Without enough data-skilled people embedded across lines of business, the central governance team cannot “move the mountain.” Recruiting, training, and embedding talent is a core part of governance success.

Collaboration and Learning Across the Industry

Xin stresses thatAI governance is still evolving as an industry. “We haven’t fully figured out what strong AI governance looks like,” she says. “That’s why collaboration is key. I connect regularly with Chief Data and AI Officers to share learnings, understand what works, and identify gaps.” 

She also emphasizes that organizations cannot wait for regulations. “Policy will eventually catch up, but responsible innovation requires us to design standards and lead with integrity today.”

Lessons for the Next Generation

For young professionals, Xin’s advice is both practical and encouraging:

• Learn Continuously: “There’s so much knowledge available for free. Study frameworks like DAMA and DCAM, attend AI conferences, and learn from how other industries handle challenges.”
• Stay Curious: “Technology will continue to evolve. Curiosity keeps you ready to adapt and innovate.”
• Balance Ethics with Innovation: “Fast innovation is useless if it erodes trust. Learn how to innovate responsibly.”

Her final reminder is powerful: “Your job is to make sure data and AI change the world for the better.”

Leadership with Responsibility

Xin’s career shows that real leadership in AI and data is not about chasing the newest tools. It’s about integrity, accountability, and the courage to slow down when everyone else wants to rush. She sees governance not as a limitation, but as the foundation that allows innovation to move faster and more safely. 

“AI and data are changing everything,” she says. “Our role is to make sure they change it for the better.” 

For Xin, the future is full of opportunity—but only for organizations willing to pair ambition with responsibility. Those that invest in clear standards, ethical practices, and continuous learning will be the ones that earn trust and stay ahead as technology evolves.

Turning Insights into Action

Xin Tu’s story demonstrates that curiosity, ethics, and practical governance can guide organizations to innovate responsibly. Her approach is a roadmap for anyone navigating AI and data:

• Start with trust—make sure your data is accurate, governed, and understood.
• Collaborate widely—across risk, technology, business, and industry peers.
• Empower your teams—build data talent, champions, and clear incentives.
• Balance risk with innovation—move fast, but with controls that people actually use.

As Xin says, “AI and data are powerful tools. Our responsibility is to make sure they are used for the betterment of organizations and society. That’s the true meaning of leadership in the digital age.”

For more stories of leaders shaping the future of data and AI, stay tuned withThe Executive Outlook.