Boards Want AI But Don’t Know How to Control It
Imagine your company deploying an agentic AI system that autonomously negotiates vendor contracts, adjusts pricing in real-time, and makes hiring recommendations. Your CDO is thrilled. Your board? They’re still figuring out what it means to govern AI agents.
This governance crisis is far from fiction and is unfolding across Asia, according to new research from the Diligent Institute, published in collaboration with the Singapore Institute of Directors (SID) and Governance Institute of Australia (GIA). Nearly half of organizations in the region (48%) are prioritizing AI adoption as their top strategic initiative for 2026. That’s higher than pursuing growth opportunities, managing cyber risks, or navigating geopolitical chaos.
The problem is that they’re building the plane while flying it blindfolded.
57% of Asian organizations have already woven AI into their operations, with 70% placing digital transformation at the top of their board agendas. Yet when it comes to governing these systems, there’s a growing chasm between ambition and capability. Only 31% have mandated director training on AI. Just 28% have recruited directors with actual AI expertise. Nearly 70% identify digital technology skills as critical board needs, yet here they are, racing ahead anyway.
“In the era of AI, the greatest risk isn’t the technology itself, but the governance gap that it is creating,” said Dottie Schindlinger, executive director of the Diligent Institute. “By developing strong expertise and robust oversight, organisations can secure a competitive advantage.”
This becomes a major problem with agentic AI, which not only assists but also acts independently. 87% of respondents see productivity gains as the upside. But 64% flag data quality and privacy concerns as top risks, and 61% admit they lack governance processes to guide AI decision-making. In other words, everyone loves efficiency and speed, but they have no guardrails.
For chief data officers navigating this landscape, the disconnect is visceral. You’re dealing with board members who think generative AI is just “fancy autocomplete” while your data pipelines feed autonomous systems making million-dollar decisions. You’re fielding questions about ROI when you should be discussing decision provenance and model drift. It’s like explaining quantum physics to someone still struggling with long division, except the stakes are your company’s future.
Organizations are scrambling to catch up. One-third are creating AI committees or working groups. 37% now require CTOs or CIOs in board meetings for AI discussions. 72% want more strategic planning time; 53% need exposure to external experts. But here’s the thing: Committees and consultants won’t close a capability gap this fundamental.
“In today’s AI-driven business landscape, corporate governance has become a critical business imperative,” said Terence Quek, SID’s chief executive officer. “Boards must prioritise director education and sustained capability development to build the resilience needed to thrive amidst increasing technological complexity.”
When nearly 70% identify digital technology skills as critical board needs, yet only three in ten mandate training, you’re ignoring an existential problem. The market has spoken: Companies are choosing innovation over caution. But as any CDO knows, moving fast without governance is recklessness with better branding.
The question now is whether organizations can afford not to upskill their teams on AI before their autonomous systems make a decision that can’t be undone.
Because somewhere right now, an agentic AI is processing data, making decisions, and shaping outcomes. And the people theoretically overseeing it? They’re still reading the manual.
Image credit: iStockphoto/SergeyNivens