The CTO is not disappearing. In a March 16, 2026, opinion article for CIO, Omilia CTO Marios Fakiolas argues that the role should shift from personally approving every technical choice to designing the systems and practices that let teams—and AI tools—make sound decisions at scale. The headline is a provocation, not a report that companies are eliminating CTOs. The useful question is whether leaders can delegate more execution without delegating accountability.
What Fakiolas means by “the CTO is dead”
Fakiolas’s target is the CTO as technical gatekeeper: the person who must inspect each architecture choice, requirement, or approval before work can proceed. As AI tools expand the volume of work teams can attempt, he argues, that model risks turning the executive into a bottleneck. His alternative is a leader who designs decision rules, review systems, and feedback loops so more work can move without relying on one person’s judgment at every step.
He captures the change this way: “The technology gatekeeping role is dying, but that doesn’t mean the CTO’s responsibilities are shrinking.” The point is not less leadership, but leadership applied to the system that produces decisions rather than to each decision individually. [CIO opinion article]
This is a normative argument, not proof that every company should remove specialist teams or that AI can reliably produce production-ready architecture. The right balance depends on the consequences of error, regulatory duties, technical maturity, and whether a team can validate the work it delegates.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitches#1 Best Overall
Why the shift is persuasive—and where the evidence stops
There is a real distinction between more AI use and more business value. McKinsey’s 2026 survey found that 80 percent of respondents said AI improved their individual productivity, while 37 percent reported some enterprise-level EBIT impact; 6 percent met McKinsey’s definition of AI high performers. These are survey responses, not causal proof that AI caused the reported outcomes. The gap suggests that productivity gains do not automatically become measurable company-wide financial results. [McKinsey, 2026]
McKinsey’s 2025 survey likewise identified workflow redesign and transformation practices among the small group it classified as AI high performers—about 6 percent of respondents. That association supports examining how work is organized, not assuming that a particular reorganization will succeed everywhere. [McKinsey, 2025]
Expectations for AI agents are also not the same as realized adoption. Gartner forecast in an August 26, 2025 release, updated September 5, that 40 percent of enterprise applications would include task-specific AI agents by the end of 2026, up from less than 5 percent at publication. That was a forecast, not a measured end-of-2026 result. [Gartner forecast]
Meanwhile, Gartner’s 2026 CIO agenda material reported that 48 percent of digital initiatives met or exceeded business targets, and 94 percent of surveyed CIOs expected major changes to plans and outcomes within 24 months. The figures underline the challenge of turning technology programs into intended results; they do not show that AI itself resolves that challenge. [Gartner CIO agenda]
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →What changes in the CTO’s work
| From | Toward | What still needs to be true |
|---|---|---|
| Personally reviewing every technical decision | Setting standards, review mechanisms, and escalation paths | High-impact decisions still receive qualified human scrutiny. |
| Choosing a supposedly permanent “right” tool or architecture | Preserving the ability to change tools or designs without unmanaged disruption | Teams understand dependencies, migration costs, and operational risks. |
| Showcasing AI demos | Targeting defined workflow outcomes such as delivery time, quality, reliability, or cost | Results are measured against a meaningful baseline, not assumed from tool use. |
| Optimizing for technical output and approvals | Measuring business results alongside quality, security, reliability, and cost | Speed does not become the only success criterion. |
| Preserving narrow handoffs by default | Testing broader end-to-end ownership where it is practical | Specialist expertise remains available where complexity or risk requires it. |
The shift is less about abolishing review than making review proportional to risk. Routine, reversible work may need lighter controls than a change affecting payments, personal data, safety, or a critical service. A CTO can define what teams may decide independently, what requires peer review, and what must be escalated—then revise those boundaries as evidence accumulates.
How to put the argument into practice
- Choose a business problem before choosing an AI tool. Define the workflow, the outcome to improve, and the baseline. A demo is not evidence of reduced cost, faster delivery, or better service.
- Set decision boundaries. Specify which work teams can approve, where human review is mandatory, and which security, privacy, reliability, or regulatory requirements cannot be waived for speed.
- Build validation into the workflow. AI-generated code, designs, or recommendations need checks appropriate to their use: tests, peer review, threat analysis, operational monitoring, or expert approval. The tool’s fluency is not validation.
- Measure outcomes and failure costs together. Track intended business results alongside defects, incidents, rework, security issues, and total cost. If speed improves while error rates or support burdens rise, the workflow has not necessarily improved.
- Adjust organization design from evidence. Broader ownership may reduce handoffs in some settings; specialist teams remain important when they provide scarce expertise or controls. Treat team boundaries as a design choice to test, not a universal obstacle.
- Keep accountability explicit. AI can help produce or assess work, but a named human leader remains responsible for the standards, risk decisions, and business outcomes the organization accepts.
The strategic role is broader, not smaller
Technology leadership is already tied to enterprise strategy in the organizations highlighted by McKinsey. Its 2026 technology-workforce article reports that two-thirds of top-performing companies had technology leaders very involved in crafting enterprise strategy, compared with 52 percent of other organizations. That is an association, not evidence that involvement alone causes stronger performance. [McKinsey, 2026]
For a CTO, strategic involvement means translating business priorities into technical choices and operating conditions: what to automate, where reliability matters more than speed, which risks are acceptable, and how the company will know whether a change paid off. Fakiolas’s second line—“The old CTO processed documents. The new CTO builds the processing systems.”—is best read as a description of that shift in leverage, not a guarantee that AI can replace expert judgment.
Who should keep the old approval model?
No company needs to eliminate executive review simply because AI tools are available. A high-control environment may reasonably require more centralized oversight when decisions are hard to reverse, errors could cause substantial harm, or laws and contracts impose specific obligations. Conversely, a mature team working on low-risk, reversible tasks may gain little from routing every choice through the CTO.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchPC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11The practical test is whether a decision can be delegated safely: Is its owner qualified? Can the result be checked? Is the change reversible? Are security and compliance requirements clear? What is the cost of a mistake? Where answers are uncertain or the downside is high, keep stronger human review and escalation. Where they are clear and the risk is bounded, design a process that lets work proceed without unnecessary executive approval.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
