Saks Global Holdings, the house of some of the most iconic luxury brands in the world, is reshaping its data and AI strategy amid seismic shifts in retail and technology. With a new unified structure and an expanding vision for intelligent, customer-centric experiences, the company is investing deeply in the foundations needed to enable responsible innovation at scale.

In the first part of this interview, Veronika Durgin, Vice President of Data at Saks, laid out her “AI readiness layer cake” framework, emphasizing the need for strong data foundations, contextual understanding, and cross-functional alignment.

In the second, she explored data-informed decision-making, guardrails for agentic AI, and the future of data talent in a rapidly evolving ecosystem.

In this final installment, Durgin and Gautam Singh, Head of Analytics, Data, and AI at WNS Analytics, examine one of the most nuanced challenges facing modern enterprises: balancing data democracy with governance, unlocking innovation safely, and preparing for the next wave of AI-powered transformation.

The push and pull of data democracy

When asked how enterprises should approach governance while enabling AI to deliver real value, Durgin describes an internal tension that every modern data leader will recognize.

“I have the angel and the devil on my shoulders,” she says. On one side, she’s a strong believer in data democracy, giving people access beyond predefined role-based slices, because they often spark meaningful innovations. “Innovation happens when we step outside of our bubble and just start making connections.”

But the other side is equally real. Unlimited access brings risk: misused data, accidental exposure, or leakage amplified by increasingly powerful AI tools.

“If you talk to anyone in security, they just want to lock everything down because it’s safe,” she notes. “If nobody has access to it, then it’s super safe.”

The challenge becomes creating an environment where both people and tools can access what they need, securely and intentionally.

Durgin points out that AI heightens this tension. If an internal LLM accidentally trains on data it shouldn’t have seen, there’s no easy fix. Even human behavior complicates things. “We as humans are not always good at protecting things,” she says, especially now that AI-powered search and copilot features make sensitive information easier to discover.

Her guiding ethos is simple: “If you treat data as your own, it gives you this ownership of what is the safe way to allow people to use it.”

The front door analogy: Governance as everyday security

Durgin turns to a home security analogy to illustrate governance: “You have a house, your security compliance is your front door. You want to make sure you have it right, but you also want to make sure that everybody who needs the key has it.”

Family, friends, and trusted people should enter. Random strangers should not. Balancing openness and safety, she suggests, is the heart of modern governance.

Creating safe environments to experiment

Speaking of how organizations can balance data democracy with innovation, Durgin mentions safe experimentation zones. “Maybe you need to hypothetically make a copy of a subset of your production data and put it in a contained environment where it’s safe for you to just do whatever you need with it,” she explains.

She compares this to a lab — a controlled environment where risk is minimized, but creativity isn’t. To reinforce the concept, Durgin invokes Amazon’s “two-way door” philosophy:

“Make sure that when you experiment and innovate, it’s a two-way door. When you still try and experiment, you can always back out of it.”

Once a system is stable, validated, and safe, that door may become one-way—but early innovation must remain reversible. At the core, she says, it is always about mitigating blast radius.

Transformation a few years ahead

When asked about the most transformative AI shift coming in the next few years, Durgin says, “I don’t know what’s going to happen in two to three years. I can barely guess what’ll happen in six months.” But she does envision broad themes.

1. Strong data foundations remain the prerequisite for everything that comes next: “If your data foundation is solid, whatever the next new thing, it’s fine.”
2. The industry finally automates the “boring things,” the routine, copy-paste tasks that burden knowledge workers today.
3. Advancements in medicine, like faster drug development, better diagnostics, and deeper self-understanding, improving the quality of life across the board.

Curiosity, learning, and real conversations

As the interview closes, Durgin circles back to what she believes truly underpins long-term success: “Make sure you’re solving real business problems,” she emphasizes, whether internal efficiency or customer delight.

Equally essential, she says, “Make sure you get your data foundation in order. Once your data is in good shape, you can bring any innovation in.”

Finally, she underscores the importance of a culture that stays curious, experiments responsibly, and stays human. “We still need to talk to each other, discuss topics and hard problems. We need to get together and solve those hard problems.”

Her message is clear: AI may advance rapidly, but the future will still be built by people who understand context, collaborate openly, and never stop learning.

Disclaimer: The interviewee’s insights are personal and not representative of any current or past employer.

CDO Magazine appreciates Veronika Durgin for sharing her insights with our global community.