When Technology Becomes Strategy: Rethinking the CIO's Mandate

There is a shift underway in how companies are run, and it is one of the more consequential of the decade. The chief information officer, for years the executive associated with keeping systems running rather than deciding where the business should go, is moving into the centre of enterprise strategy.

McKinsey's Global Tech Agenda 2026 puts a figure on it: nearly two-thirds of top-performing companies now say their technology leaders are "very involved" in shaping enterprise strategy, compared with 52% of everyone else. Gartner's research describes the same migration from a different vantage point.

For now, the shorthand the industry has settled on, technology is strategy now, is, for once, more than a slogan. The shift is real and worth taking seriously. It is also only the opening chapter, and the more interesting part of the story is what happens after the title changes.

In the same period that surveys were describing CIOs as strategy architects, MIT's NANDA initiative published The GenAI Divide, which found that roughly 95% of enterprise generative-AI pilots delivered no measurable impact on the profit-and-loss statement, with only about 5% producing real return.

Gartner's 2026 CIO and Technology Executive Survey, drawing on more than 2,500 leaders, is consistent with that picture: only 48% of digital initiatives meet or exceed their business targets, and the share of companies stepping back from most of their AI efforts has risen sharply year over year.

Both things are true at once. Technology leaders are being welcomed into the strategy conversation, and a great many AI programs are not yet delivering. That tension is not a contradiction to explain away; it is an accurate description of where enterprise leadership stands in 2026, and understanding it is what separates the companies that will pull ahead from the ones that will spend the next two years busy without becoming better.

What genuinely changed, and it did change

The structural shift deserves credit before we examine its limits.

For most of the last two decades, technology lived downstream of strategy. The business decided where it wanted to go; technology was handed a budget and a deadline and asked to wire it up. That sequence made sense when technology was a cost of doing business; infrastructure you maintained so the real work could happen on top of it. It made far less sense once the product itself became the technology, the distribution channel became the technology, and the cost structure became the technology. It is difficult to keep downstream the very thing that now determines what a company can offer.

The data captures this turn clearly. McKinsey reports that the number of organisations continuously co-creating business and technology strategies throughout the year has nearly doubled. Among the highest-performing companies, nearly half have adopted this collaborative approach.

Gartner frames the same idea from the planning side: only 18% of CIOs have adopted dynamic, off-cycle reprioritisation, yet those who have are 24% more likely to be top performers. The annual technology plan, that artefact of the budgeting calendar, is quietly giving way to a rolling dialogue between what the business wants and what technology can now do; a conversation that never fully closes.

This part of the story is solid. The centre of gravity has moved. Technology expertise has become, in McKinsey's phrasing, a form of strategy expertise, and the organisations setting the pace treat their CIO as a co-author of where the enterprise is heading rather than as a service provider working to a service-level agreement.

The opportunity, then, is genuine and large. What deserves equal attention is the work that turns the opportunity into outcomes, because the title and the capability do not arrive together.

A seat at the table reflects value already created

Here is the idea I would put at the centre of this discussion. Being invited into the strategy conversation tends to follow the creation of value rather than cause it. By the time a CEO is asking the CIO what the company should do, much of the work that determines whether the answer is any good has already happened, or already stalled, somewhere in the operating model. A seat at the table is most often earned by demonstrating that technology can be converted into results. Organisations that grant the seat in the hope that proximity to leadership will, on its own, produce those results may be reading the sequence in reverse.

The 95% figure is what gives this weight. If elevating the CIO were sufficient on its own, more pilots would be working, because the surveys say the elevation has largely happened. MIT's researchers were precise about the reasons they are not, and the finding is worth carrying into every boardroom that believes an org-chart change has settled the matter: the cause of failure is overwhelmingly organisational, not technological. They called it a learning gap: the difficulty of folding AI into real workflows, structures, and culture. The model was rarely the obstacle. The model was demo-ready. The enterprise around it was not.

This is the distinction between a strategic mandate that exists on paper and one that shows up in results. The first arrives at the meeting with a roadmap of impressive-sounding initiatives. The second arrives having already rebuilt the machinery that turns initiatives into P&L impact, and often has fewer programs to point to, because the discipline of scaling one thing properly tends to matter more than the appearance of piloting ten.

There is a quieter point in the MIT data that is easy to pass over. The study found that more than half of generative-AI budgets were directed at sales and marketing, while the strongest returns were sitting in less glamorous back-office work: document processing, compliance, and internal operations. The money flowed to where AI was most visible, not where it was most valuable. That is an understandable instinct, and an expensive one, and avoiding it is one of the clearest markers of a technology agenda set with discipline.

What do the leading companies do differently?

Three things stand out, and none of them is buying a better model.

They change the operating model before they change the tooling. McKinsey's standout examples are instructive precisely because they are unglamorous in the right way. DBS Bank, now regularly cited among the world's best digital banks, did not arrive there by procuring more AI. It reorganised the institution into more than thirty customer- and capability-aligned platforms, each jointly led by a business leader and a technology leader. The purpose of that restructuring was to remove handoffs. When decisions that once took months take days because the people who understand the customer and the people who understand the system sit on the same team, you have built the foundation that lets AI investment pay off. The AI is the more straightforward part. The operating model is the durable advantage.

They build the data foundation that makes autonomy safe. Agentic AI, systems that plan, decide, and act across a workflow rather than simply answering a prompt, is the ambition almost everyone now shares. Yet McKinsey found that a quarter of even top-performing companies acknowledge they lack the data foundations to scale it securely and reliably. Agents are only as trustworthy as the data, and the guardrails beneath them, and in regulated industries- banking, insurance, and lending- the cost of an agent acting on poor data is not an awkward demo but a compliance event. The leaders treat data quality and governance as a precondition rather than a task to be cleaned up later.

They build capability in-house rather than only renting it. This is the lever I find most under-discussed and most revealing. McKinsey's top performers plan to increase insourcing, bringing strategic technology expertise back in-house at a notably higher rate than their peers, and they are reskilling their own people rather than relying primarily on vendors and outsourced teams. The logic is clean: outsourcing buys capacity, but insourcing builds capability. MIT's data points in the same direction from another angle; it found that buying AI tools through vendor partnerships succeeded roughly twice as often as internal builds. Read together, the lesson is not "build nothing." It is closer to "buy the model, own the integration." The capability that connects a tool to the actual business is the part that should stay in-house, because that connection is where the value is created.

The common thread is worth naming. Aviva, the UK insurer, deployed more than eighty AI models across its claims journey and cut liability-assessment time by twenty-three days, while improving routing accuracy by 30% and reducing customer complaints by 65%. It is a remarkable result, and, as McKinsey is careful to note, it came alongside a full operating model and cultural transformation, not in place of one. The eighty models are the headline. The operating model work is the reason the headline is true.

The geography that can no longer be averaged away

There is a dimension the global surveys tend to flatten and that, in my view, matters a great deal: where a company sits changes what the job actually is.

Gartner surfaced one of the sharpest fault lines in recent memory: a clear divide between US and non-US CIOs on vendor strategy. Roughly half of CIOs outside the United States expect to change how they engage vendors based on regional and geopolitical factors, against around 31% inside it, and twice as many non-US CIOs are deliberately shifting toward regional providers. Gartner's chief of research went so far as to describe it as the possible beginning of a shift in technology hegemony. In plain terms: data sovereignty and vendor geography have moved from procurement footnotes to board-level strategy inputs. For a CIO in Frankfurt or Mumbai, where the AI stack is hosted, it is now a strategic question, not an operational one.

India is the case that resists the simple narratives in both directions, and it rewards a closer look because it shows how uneven this transition really is. On one hand, the country is a genuine AI heavyweight; ranked first globally in AI skill penetration, second only to the US in weekly active users of the leading consumer AI tools, and the site of sovereign-compute commitments running into the tens of billions of dollars alongside a wave of enterprise partnerships: Infosys tying up with both OpenAI and Anthropic for agentic deployments in regulated sectors, and L&T and NVIDIA building "sovereign-by-design" AI factories. On the other hand, the same research that celebrates India's scale is candid that scale is not the same as depth. Surveys of Indian enterprises show a clear majority still running pilots, with fewer than a quarter having moved AI into mission-critical production and a large group of early adopters still in the experimentation phase. The talent and the consumer adoption are world-class; the enterprise-grade discipline to convert that into production systems is, as it is nearly everywhere, still being built.

The reason the geographic lens matters is that it complicates the simplest version of the CIO-as-strategist story; the one that imagines a single global playbook. There isn't one. The US technology leader is optimising for speed and scale within a relatively settled vendor landscape. The European leader is threading sovereignty, the EU AI Act, and a fragmenting vendor map. The Indian leader is operating amid extraordinary talent density and a still-maturing production muscle. The strategic question of how technology creates competitive advantage here has a different answer in each, and the most effective leaders adapt the playbook to the context rather than importing one wholesale.

What the genuinely strategic CIO is focused on in 2026

If the title alone is not enough, and the operating model is the thing, what does the work look like in practice? Here is my read, and it asks more than the standard checklist.

First, the strategic CIO moves the organization off the annual cycle. Not because continuous planning is fashionable, but because an annual cadence structurally guarantees that technology decisions lag business reality by up to twelve months; a long time when the underlying capability changes every quarter. Gartner's finding that off-cycle reprioritisers are meaningfully more likely to be top performers reads less like a curiosity than a prompt to act. Quarterly business-technology reviews are a sensible start, working toward dissolving the line between the two planning processes altogether.

Second, the strategic CIO works fluently in the language of the P&L. This is the shift that most clearly distinguishes a mandate that lands from one that does not. Gartner found that CIOs who consistently pursue financial outcomes from technology, especially AI, are 25% more likely to excel, yet only about a third do so consistently. The opportunity is in drawing a straight, defensible line from a deployment to a number on the income statement, and in bringing that line, rather than uptime and ticket-resolution metrics, into the room where strategy is set.

Third, and this is where I would place the most emphasis, the strategic CIO treats change management as a core technical challenge rather than a soft adjunct to be delegated. MIT's learning gap, and McKinsey's finding that top performers cite change management as a larger obstacle than their peers do, in part because they have progressed far enough to encounter it, point in the same direction. The binding constraint is rarely the model. It is whether thousands of people will actually change how they work. The leaders who do best in 2026 internalise that the hardest part of an AI transformation is not the AI. It is the transformation.

The complete version of the story

The encouraging narrative- technology leaders ascend to strategy, technology becomes the growth engine- is not wrong. It is simply the first half. The fuller version is this: technology has become inseparable from strategy, which means the opportunity for technology leaders has never been greater, and the distance between those who can convert that opportunity and those who hold the title is widening just as quickly. McKinsey describes it as a widening maturity gap. MIT calls it a divide. Whatever the label, it is real, and it is hardening into lasting advantage for the companies on the right side of it.

So I would offer a constructive note to any organization feeling good about having brought its CIO into the strategy conversation this year: that is the right move, and it is a beginning rather than a conclusion. The structural shift the surveys describe is a starting line. The leaders who will define 2026 are not only the ones invited into the strategy room.

They are the ones who, over time, make the strategy conversation and the technology conversation the same conversation; and then do the patient work of rebuilding the organisation so that the decisions made there survive contact with reality.

That is a harder undertaking than securing the invitation. It is also the one that ultimately counts.