Two-thirds of CIOs are accountable for AI they don't control
Powered by Single Fin
Welcome to this week’s edition of CIOsurge!
This week:
Woo An, CEO of Zaun, on why you cannot govern what you have never built yourself
A new IBM study: two-thirds of CIOs are accountable for AI systems they do not fully control
Info-Tech LIVE 2026: AI execution is pushing CIOs back to IT fundamentals
Let’s make this week a game-changer.
Stay sharp. Stay ahead.
Build It Yourself Before You Buy It: A CIO's AI Adoption Test
One of the best pieces of advice from my conversation with Woo An, CEO of Zaun, applies whether you run a security team or sit in the CIO chair: go build something yourself.
I asked Woo how a CIO should approach AI adoption when the space changes week by week. He started with the build-versus-buy question, because it is the one every leader is wrestling with. Teams are going to build a lot of custom tooling and buy a lot of AI products at the same time, and the CIO has to understand the tradeoff. "You need to understand the pros and cons of building your own tool and enforcing your own security control based on your compliance, versus buying something where you have less control if it's a SaaS product." The best case is a SaaS tool that already meets every control you need, but Woo was blunt that this is extremely rare right now, close to impossible. His advice was to stay on top of it and figure out when the ROI flips from build to buy.
Then he gave the answer that stuck with me. "Whether you're a CIO looking to get into AI, or a new grad in computer science looking for a job, just go out and build projects. Use these AI tools and build a website, build a tool, build something. You have to feel it. You have to understand how people are going to use this. Then you can dig into the mechanisms of how it works, where the gaps are, and where the signals are actually useful."
I could not agree more, because it mirrors my own experience. As I built out our agent infrastructure at SingleFin, now dozens of agents running internally, it changed my lens on the world completely. But the moment I started deploying agents at scale, I saw exactly how many places security could become a serious problem. You do not see that from a slide deck. You see it from building.
Woo made the same point about leadership credibility. "If you don't use AI yourself and understand why you're enforcing it, the rest of the org won't adopt it. There's a fakeness that shows if you tell people to do something and you don't do it yourself."
The takeaway for CIOs and CISOs: you cannot govern what you have never used. Build a small AI project this month. The gaps you find by hand will shape your strategy better than any vendor briefing.
If you want to discuss our AI development path: Book time with Zack Tembi: 30 minutes meeting with Zack Tembi
Two-Thirds of CIOs Are Accountable for AI They Do Not Fully Control
A new IBM Institute for Business Value study of 2,000 C-level technology executives found that as AI moves from experimentation to enterprise-wide deployment, two-thirds of surveyed CIOs and CTOs report being held accountable for AI systems they do not fully control. The lack of visibility is widespread: 70% say teams across the business are deploying technology faster than IT can track. At the same time, technology leaders face growing pressure to scale AI faster, even as many lack the governance structures to support it.
This is the number that names the problem I hear every week. The accountability has already landed on your desk. The control has not. When 70% of leaders admit the business is shipping AI faster than IT can even see it, you do not have a technology gap, you have a governance gap. Start with an inventory of what is already running, then build the approval and monitoring layer around it. You cannot be accountable for a system you cannot see, and right now most CIOs are signing up for exactly that.
- Zack Tembi
AI Execution Is Pushing CIOs Back to IT Fundamentals
Info-Tech Research Group's Best of 2026 Mid-Year Report finds that AI has moved from a strategic ambition to an execution challenge, and CIOs are responding by concentrating on the foundations required to scale it with confidence. The report, drawn from the firm's most-accessed research across the first half of 2026, shows strong demand for guidance on AI execution, cybersecurity, data, infrastructure, governance, risk, and workforce readiness. The throughline from Info-Tech LIVE 2026 was direct: AI value depends on disciplined execution, strong governance, and reliable data foundations, not on the newest model.
The industry spent two years treating AI as something separate from the IT discipline that came before it. The correction is here, and it is healthy. Clean data, access controls, change management, and a real governance framework are the boring work that determines whether AI delivers or stalls. If you built those foundations over the last decade, this is your moment. If you skipped them to chase pilots, the bill is coming due. The fundamentals were never optional, and AI just made that obvious.
- Zack Tembi
💡 CIO Spotlights
Chaim Mazal named Chief Information Security Officer at GitLab
Joins GitLab to lead the company's global security organization, covering both the business and the platform
Comes from Gigamon, where he served as Chief AI and Security Officer overseeing security and AI governance
Will help shape GitLab's approach to AI security as enterprises adopt AI agents and face emerging AI-driven threats
A former GitLab customer, brought in for that practitioner instinct alongside his security depth
