"If You're Not Uncomfortable, You're Not Governing AI"

🎯 This article is part of my Board Evolution: A Tech Leadership Interview Series

"Nobody wants to look stupid."

That's what Kobi Leins (PhD, GAICD) told me when we sat down to discuss why boards struggle with AI oversight. Kobi doesn't just talk AI governance—she's knee-deep in it daily. With a background spanning UN advisory roles, academia, and commercial implementation, she's the person organizations call when the AI hype collides with operational reality. As she puts it:

"The biggest risk isn't asking a naive question—it's assuming the strategy is sound because it sounds impressive."

Where Boards Are Fumbling the AI Ball

Too often, boards are falling into three traps:

• They assume someone else understands it. Kobi sees too many directors leaning on surface-level briefings or a single tech advisor—without pressing further. When AI becomes a buzzword that isn’t underpinned by sound understanding, decisions lose rigour and, ultimately, impact.
• They over-index on skills, underinvest in culture. Boards are scrambling to hire "AI experts." But unless those people are empowered to challenge assumptions and bridge across disciplines—legal, product, risk, ethics—the expertise stays siloed and underused. "You need someone who can speak data science and translate to legal and ops. That's not a unicorn—it's just a different way of hiring."
• They treat governance as a blocker. "It's the opposite," Kobi says. "If you have guardrails, you move faster." Trying to retrofit policies after implementation usually creates more friction than momentum.

What the Most Effective Boards Are Doing

Unsurprisingly, the boards getting AI right are doing the work that matters.

✔️ Starting with risk appetite. They begin with what they're comfortable with—and align reporting, oversight and incentives accordingly.

✔️ Embedding AI in strategic dashboards. Rather than pushing AI to a subcommittee, they treat it like cybersecurity, customer experience or ESG: core to how the business operates and grows.

✔️ Incentivising ethical choices. They don't just publish values—they back them up. Bonuses, reviews and codes of conduct all reflect AI-related responsibilities.

"You can't operationalise ethics without accountability. The principles have to mean something."

Boards vs Execs: Where the Gaps Show

When execs own the tech stack and implementation details, boards can mistakenly think their job is just to nod along.

"If you're not asking what data was used, or how decisions are being made, you're not governing—you're observing."

Kobi warns that AI systems bake in strategic decisions. Whether it's risk thresholds, pricing logic or customer profiling, these aren't just technical tools—they're reflections of company intent.

When Boards Just Watch From the Sidelines

The problem isn't theoretical anymore. From the robo-debt fiasco to biased hiring algorithms and privacy-crushing marketing tools, we're watching real organizations face real consequences. 

"I've seen simple automation cause serious damage, and highly complex AI tools do very little. The key is context, scrutiny, and challenge—at every level."

And yet, many organisations are still stuck in paralysis:

"Boards come to me and say, 'We're so embarrassed—we haven't done anything yet.' But that first conversation? That's the turning point. The rest follows quickly."

Culture Over Credentials

"You don't need to be an expert. But you do need to feel comfortable asking uncomfortable questions."

This is where board culture matters most. Kobi points to the example of kindergarteners vs CEOs solving a problem: the children solved it faster—not because they were smarter, but because they were less constrained by hierarchy and ego.

That same playfulness and curiosity is exactly what boardrooms need more of.

Board Checklist: What Boards Should Be Doing Now

• Define your AI risk appetite Align with strategy—and review it regularly.
• Make AI a standing agenda item Like cybersecurity or ESG, it deserves ongoing scrutiny.
• Integrate AI into your reporting dashboards Track usage, risks and opportunities systematically.
• Tie principles to incentives Link ethical standards to performance metrics.
• Challenge assumptions Ask specific questions about data, design and oversight.
• Upskill the whole board Don't rely on a lone "tech-savvy" director.
• Foster a culture of curiosity Normalise not knowing. Reward thoughtful questioning.

Exec Checklist: What the C-Suite Should Be Getting Right

• Map what's already in play Document all AI and automation in use—by team and use case.
• Apply AI to real business problems Focus on known gaps—not shiny distractions.
• Involve legal, risk, and ethics from the start Bake governance into the build process—not as an afterthought.
• Hire translators, not just technologists Bridge technical and strategic conversations across the business.
• Track the environmental footprint Include AI in ESG frameworks—especially around water and energy use.
• Develop clear narratives Build the capability to explain AI decisions to boards, customers, and regulators.

Final Thought

"You don't need all the answers. But you do need to start asking better questions."

AI governance isn't about getting everything right on day one. It's about creating enough momentum to keep learning. The organizations pulling ahead aren't showcasing fancy tech demos—they're building cultures where candid questions trump polished presentations, where clarity beats complexity, and where governance enables rather than restricts.

Thanks to Kobi Leins for the clarity, honesty and insight in this conversation.

If this piece sparked thoughts—or discomfort—that's exactly the point.