The title of the Chief Information Officer is something of a running joke in executive circles, with the role’s acronym variously interpreted as: “Career Is Over,” “Constantly In Overload,” or “Chief In Trouble Officer.” Humor aside, the CIO role has grown to be one of the most demanding, most scrutinized, and most frequently turned over in the C-suite, a function that rarely receives the strategic recognition its complexity deserves.
Artificial intelligence is not simply adding to that pressure; it is reframing the question entirely. Now, the core of what a CIO does – managing information, overseeing systems, and enabling access to technology across the organization – can increasingly be done by, automated by, or meaningfully augmented by AI. More disruptive still, AI is placing capabilities that once required specialist IT knowledge directly into the hands of ordinary users across the organization. When anyone can build an application, query a dataset, or deploy an automated workflow without writing a line of code, the organizational rationale for a traditional IT function comes into question in ways that would have seemed implausible just a few years ago. As Vikram Nafde, EVP and CIO at Webster Bank, has
Talk to CIOs who have lived through multiple technology cycles, and you will hear something striking: many say they have never been less certain about what their role actually requires, what skills it demands, or what scope the function should have. The landscape is moving faster than any planning cycle can absorb, and leaders are clear-eyed enough to say so. This article presents that landscape and offers a practical guide for CIOs navigating one of the most significant transformations their role and their function have faced.
The CIO role has already transformed across three distinct eras. In the 1990s and early 2000s, the CIO was fundamentally an infrastructure manager. The mandate was operational: keep systems running, provide employees with reliable access to technology, manage data centers, and deliver basic IT support. Technology was a back-office function, and the CIO was its caretaker.
The rise of e-business and cloud computing forced a first reinvention. As organizations extended beyond their own walls, integrating external partners, platforms, and ecosystems, enterprise architecture emerged as a defining CIO competency. Connecting systems intelligently, managing APIs, and governing data that now lived outside the organizational perimeter became central concerns. Cybersecurity, once a niche specialty, became a board-level topic almost overnight.
The era of digitization and big data demanded a second reinvention, this time toward business-centricity. CIOs could no longer afford to think in purely technical terms. Understanding business processes, identifying where technology could unlock transformation, and contributing meaningfully to long-term strategy became key expectations. Increasingly, the CIO warranted a seat at the executive table, not as the IT expert in the room, but as a strategic partner.
Now, AI is catalyzing a third and far more radical shift. Coding is becoming a mainstream capability. Large language models (LLMs) and generative AI are placing creative and analytical power in the hands of every employee. Users are becoming ‘citizen developers’ and ‘vibe coders’. Meanwhile, CIOs must simultaneously introduce agentic AI frameworks, establish observability and guardrails, manage the economics of LLM token consumption, and mount an increasingly sophisticated defense against cyber threats, all while maintaining the operational excellence and user service that remain non-negotiable.
The CIO has evolved from specialist to orchestrator, from IT manager to business executive: the Chief Information Officer is becoming, in every meaningful sense, a Chief Intelligence Officer.
Design thinking, long a staple of innovation labs, is poised for a second peak as the discipline that connects user needs, business outcomes, and AI capability in a practical and human way.
Beneath the strategic shifts reshaping the CIO role, several powerful forces are simultaneously disrupting the technology landscape itself. CIOs who fail to anticipate them risk building on foundations that are already eroding.
The first concerns the traditional Software as a Service (SaaS) model, and whether it has a long-term future in its current form. For two decades, SaaS was the dominant paradigm for enterprise technology: standardized applications, subscription pricing, and regular updates. That model is under significant pressure, though the picture is more nuanced than a simple story of obsolescence. As AI lowers the barrier to custom software development, individuals and departments can increasingly build tailored tools without any coding expertise. This is already creating deflationary pressure on software pricing, and vendors are responding by embedding AI agents into their platforms. But the more important shift is architectural: organizations will increasingly build AI-driven interfaces and workflows on top of existing systems of record rather than replacing them wholesale. The result is not a world with less software. It is a world with more software, structured differently, and considerably harder to govern.
The second force is the rise of data as the organization’s most strategic asset. If AI is the engine, data is the fuel, and the quality of that fuel determines everything. Data products, data catalogs, and FAIR (Findable, Accessible, Interoperable, Reusable, and increasingly, legally compliant) principles are moving from technical best practices to boardroom priorities. Organizations that have invested in clean, well-governed, accessible data will be able to move fast. Those that have not will find their AI ambitions bottlenecked by the very asset they have been generating for years without properly managing. For the CIO, this means elevating data governance from a compliance exercise to a core strategic program.
Most organizations will never achieve perfect data quality, and waiting for a complete, pristine data estate before committing to AI is a mistake. The real challenge is building AI systems and decision processes that perform robustly on imperfect data and developing organizational maturity around managing uncertainty rather than demanding standards from AI that human processes never consistently achieved, either. A meeting summary produced by AI may be imperfect. So, it should be noted, was the one produced by a human.
Closely related is a third force: knowledge management. As AI systems become the connective tissue of the enterprise, the quality of organizational knowledge, how well it is structured, linked, and accessible, becomes a foundational dependency. The ability of an AI system to reason well about a business problem depends directly on the richness of the knowledge it can draw upon. Knowledge management, long sidelined as a discipline after its brief prominence in the early 2000s, is due a significant revival. CIOs who invest now in knowledge graphs, semantic layers, and structured content repositories will find that their AI initiatives land faster and with greater depth of impact.
The fourth force is the rapid evolution of how AI is accessed and deployed across the organization. The early instinct of giving everyone an AI assistant and calling it a strategy is already giving way to something more complex. Enterprises are moving toward multi-model environments, selectively combining LLMs and small language models from providers such as Anthropic, Google, and Perplexity depending on the task, the privacy requirements, and the cost profile. Modalities are expanding beyond text to voice, image, and video. Agentic AI – systems that can reason, plan, and act autonomously across workflows – introduces an entirely new architectural layer, encompassing both third-party agents and internally built ones.
The CIO’s task is to design an AI architecture that accommodates this complexity today while remaining genuinely flexible for what comes next. The key words here are model optionality and architecture modularity. One note of caution: the large unifying AI platform plays are proving slower to materialize than anticipated. CIOs should not wait for the perfect architecture before acting. Build modularly, ensure that individual components can be replaced as the landscape evolves, and treat replaceability as a deliberate architectural principle rather than an afterthought. The AI era has not finished surprising us.
Transform your career and business with next-generation AI and digital skills
The IT function must change in response. Many CIOs have already begun experimenting with different models for embedding data and AI into their organizations, and no single operating model has yet emerged as the clear winner. In some organizations, AI capability is centralized within IT; in others, it is distributed across business functions; in many, it is a hybrid of both. What matters most is finding a setup that works within a given organization’s culture and structure. But regardless of the model chosen, one principle holds: the CIO must remain the guardian of the AI architecture. Governance, coherence, and strategic direction cannot be delegated away.
There is a persistent myth worth dispelling: that AI simplifies the CIO’s operational reality. It does not. More AI means more software, more integrations, more vendors, and more architectural decisions to make and stand behind. The surface area the CIO must govern is growing, not shrinking. Cost discipline is becoming a critical new competency, particularly around model consumption, cloud infrastructure scaling, and the hidden costs of moving AI pilots into production at scale. CIOs who treat AI purely as an innovation agenda without an equally rigorous cost management discipline will find themselves with difficult conversations ahead.
A more fundamental shift is also underway. Historically, every piece of enterprise software was designed around human interaction: screens, workflows, the assumption that a person would eventually be on one end of the transaction. AI-native software does not carry that assumption. Agents read, write, decide, and act without a human in the loop. This changes how CIOs think about procurement, build versus buy decisions, and integration architecture in ways most IT functions have not yet fully worked through. The question is no longer simply whether a system is usable. It is whether it is composable, auditable, and governable by both humans and machines.
Several core competencies will define the IT function going forward. Enterprise architecture becomes even more critical, as the complexity of connecting AI systems, data platforms, agents, and legacy infrastructure demands genuinely sophisticated design thinking. The data function must be elevated accordingly. And as citizen development accelerates, CIOs will need to build real no-code and low-code expertise across IT, empowering business users to build responsibly while maintaining appropriate guardrails. This is not a small ask: citizen developers need meaningful support, training, and clear boundaries to operate effectively and safely. Most importantly, they need guidance on how self-developed software can be put into production, and how it can be taken out of the environment to avoid a graveyard of unused software products.
It is also worth reframing how the IT function thinks about business-led technology development more broadly. Shadow IT has long been treated as a problem to suppress. In the AI era, multidisciplinary teams in which business and technology expertise sit alongside each other – what might more accurately be called embedded IT – are often where the most grounded and effective AI solutions emerge. The CIO’s role is not to prevent this but to ensure it happens within a framework of architectural authority: clear standards around data governance, legal exposure, security, and code review that apply consistently regardless of where in the organization the development is taking place. As AI enablement increasingly happens within business domains rather than in IT itself, the most effective IT teams will be those that combine deep AI expertise with understanding of business processes. The days of IT as a purely technical function are over. Design thinking, long a staple of innovation labs, is poised for a second peak as the discipline that connects user needs, business outcomes, and AI capability in a practical and human way.
Managing AI proliferation across a large organization also requires segmentation rather than a one-size-fits-all approach. Effective CIOs are distinguishing between different user personas: end users who need accessible, approved tooling; professional developers who need agent orchestration platforms and coding environments; and business functions investing in domain-specific AI applications. Each group needs a different governance approach, a different level of enablement, and carries a different risk profile. Applying a single framework across all three tends to produce either over-restriction or loss of meaningful oversight.
A new governance responsibility is also emerging that few IT functions are currently structured to handle: the management of machine workers. AI agents are active participants in business processes, executing tasks, making decisions, and in some cases interacting directly with customers and colleagues. IT must develop the capability to onboard, manage, audit, and retire AI agents with the same rigor they would apply to human employees. This includes not only access rights, performance standards, and behavioral guardrails, but also the legal dimension: record retention, audit trails that hold up over extended time horizons, and clear accountability for decisions made by automated systems. The CIO who builds this governance capability early will have a meaningful advantage over those still treating agents simply as software deployments.
AI will also transform how IT operates internally. The traditional software development lifecycle is giving way to what might be called an AI-driven development lifecycle. Code generation, testing, documentation, and deployment are increasingly handled by AI tools and autonomous agents. Infrastructure monitoring, incident response, and security patching are being automated from end to end. Tasks that once required entire teams, from writing boilerplate code to triaging support tickets to generating technical specifications, are now completed in minutes. The engineer’s role shifts from doing to directing, reviewing, and governing.
The scale of this shift should not be underestimated. Research by The Hackett Group found that generative AI is expected to reduce IT operating costs by as much as 37%, freeing up significant resources for more strategic work. This is not primarily a cost-cutting story. It is a reinvestment story: fewer people will be doing routine work, more people will be doing work that genuinely matters.
Innovation remains a non-negotiable mandate: continuously scanning the market, identifying where the next wave of AI capability will come from, and translating that awareness into organizational readiness is a responsibility that cannot be outsourced or automated.
The overall implication is counterintuitive: operational roles in IT will shrink significantly as AI absorbs routine activity. But architecture, integration, and governance roles will grow in both importance and likely in number, driven by the sheer complexity of the AI-enabled enterprise. The IT function will have strategic significance across the entire organization.
The CIO’s role is closer in spirit to that of a Chief Human Resources Officer - building organizational capability - than to a project owner delivering a system.
Driving transformation requires more than technical authority; it demands the ability to inspire, educate, and bring people along through significant uncertainty. The most effective CIOs will be those who can explain AI not just to engineers but to boards, executive committees, frontline managers, and understandably skeptical colleagues. Teaching and translating are as important as deciding and delivering.
The CIO must also operate as an orchestrator of an increasingly complex workforce, one that blends human and machine capability in ways organizations are still learning to manage. As AI agents become embedded in everyday workflows, the CIO is responsible for integrating digital workers alongside human ones in a way that is transparent, purposeful, and trust-building. As Troy Gerber, CTO at Fusion5, has predicted, “In the next two years, 30% of our workforce will be digital agents. They’re not going to be replacing people. They’ll be working alongside people.” Normalizing this shift, and helping the organization navigate it with confidence, is a distinctly human leadership challenge.
Boards and executive committees are hungry for guidance on AI and often uncertain about where to begin. The most effective CIOs are discovering that their most valuable contribution at this level is not to own the AI agenda; it is to ensure the right people own it. The CIO’s role is closer in spirit to that of a Chief Human Resources Officer building organizational capability than to a project owner delivering a system. That means coaching CEOs and business leaders to take genuine accountability for AI outcomes, ensuring that responsibility stays in the functions where the value is actually created, rather than drifting upward into IT. Where the CIO becomes the de facto owner of all things AI, accountability diffuses and momentum tends to stall.
The strategic partnership between the CEO and CIO is a critical success factor in the AI era. At Workiva, the CEO and CIO have publicly championed a shared vision for AI. As they wrote in a joint article for the World Economic Forum, “The CEO sets the vision and direction… The CIO applies knowledge of every part of the company to build processes that fulfil the CEO’s vision and direction.” This collaborative model ensures that AI initiatives are not just technologically sound but are also tightly aligned with the company’s strategic priorities.
That partnership also extends to the business unit AI leads. Every function increasingly has its own AI capability, its own priorities, and its own pace of adoption. The CIO must coordinate this distributed network, aligning on architecture and standards without stifling the local initiative that often produces the most interesting results. The CIO will also have to pay attention to the right level of data and AI literacy across the organization. As an example, at Fusion5, the technology leadership has made AI literacy mandatory for all employees, demonstrating a top-down commitment to building a workforce that is ready to embrace and leverage AI. They have even gone so far as to introduce their new ‘digital agent’ employees at company-wide meetings, normalizing the concept of a human-machine workforce.
The CIO’s role is to enable and govern, not to become the organization’s universal AI expert.
A note of pragmatism is also warranted here. The CIO cannot, and should not, try to absorb domain expertise from across the business. The enthusiasm of boards for all things AI is real and understandable, but it needs tempering with honest counsel about what is realistic, what will take longer than expected, and where accountability belongs. The CIO who over-promises, or who centralizes so much that IT becomes a bottleneck, will find credibility and momentum erode quickly. Domain knowledge and specialization remain essential. The CIO’s role is to enable and govern, not to become the organization’s universal AI expert.
The best CIOs are leading with transparency, making their portfolio of choices visible to the organization, and earning trust through intellectual honesty about what is known, and what remains genuinely open. The CIO who builds an organization capable of continuous recalibration will consistently outperform the one committed to a fixed roadmap.
Finally, the CIO must become a genuine ecosystem manager. The most valuable AI capabilities will rarely be built entirely in-house. Partnering with innovative technology players, engaging with the startup ecosystem, and maintaining close relationships with research institutions and platform providers will be essential to staying ahead. In the AI era, what you connect to matters as much as what you build.
Not every AI opportunity is worth pursuing, and not every promising pilot deserves to be scaled. The CIO must develop a clear-eyed view of which AI investments align with the organization’s strategic priorities, its level of data maturity, and its capacity to absorb change.
A particular challenge at this moment is what might be called the mushrooming problem. Across most large organizations, AI pilots are proliferating at speed, initiated by enthusiastic business units, individual teams, and technology vendors alike. Many of these experiments are valuable. Many are duplicative. Some are potentially risky. The CIO must find a way to channel this energy productively, creating the structures, platforms, and governance mechanisms that allow promising pilots to be identified, assessed, and scaled with repeatability, rather than left to wither in isolation or grow unchecked without oversight.
As the number of AI tools, models, and platforms continues to expand, one of the most critical CIO skills is the ability to make clear and defensible decisions under uncertainty and with incomplete information. Going beyond prioritization, this skill requires judgment, the willingness to commit to a direction, to deliberately stop certain things, and to be accountable for those calls even when the landscape may shift again within months. Consolidating AI innovation in the early phases, rather than allowing every unit to run its own experiments in isolation, can help surface how much different parts of the organization have in common, and build a shared foundation faster than a fully distributed model would allow.
Equally important, and frequently underestimated, is having a clear process for stopping pilots that are not delivering. In practice, this is harder than it sounds. AI initiatives tend to attract organizational enthusiasm, and the ambiguity of early results makes it easy to keep investing in hope rather than evidence. The CIO should establish explicit exit criteria for every pilot at the outset, defining what would need to be true at each stage for the initiative to continue, and what would trigger a decision to stop. Making that process visible and consistent across the portfolio removes the political difficulty of individual cases and builds organizational credibility for governance as a whole.
The CIO should establish a portfolio view of AI initiatives across the enterprise, applying consistent criteria around strategic fit, feasibility, data readiness, and expected business impact. This ensures that the organization’s finite energy goes toward the initiatives most likely to generate lasting value. Maintaining that portfolio view requires active transparency, which reduces duplication, builds organizational trust, and surfaces the reality of what is working and what is not.
Underpinning all of this is the need for a robust value framework. Connecting AI initiatives to tangible business outcomes, whether revenue growth, cost reduction, risk mitigation, or customer experience improvement, is what separates an organization genuinely benefiting from AI from one simply accumulating proofs of concept.
Managing AI risk is a precondition for moving fast sustainably and securely.
As AI tools proliferate across the organization, often adopted directly by business users without IT involvement, the CIO faces the challenge of shadow AI: employees sharing sensitive data with external models, using unvetted tools, or building automated workflows that nobody in IT is aware of. The consequences can range from data leakage to regulatory breach to reputational damage. The CIO must establish clear policies, accessible guidance, and practical governance mechanisms that give employees a safe and approved path to AI adoption rather than driving usage underground. This is increasingly a board-level concern: a recent BCG report found that 74% of leading AI companies are now continuously monitoring compliance with their responsible AI frameworks.
Alongside shadow AI, the CIO must champion the principles of responsible AI. Algorithmic bias, a lack of model transparency, and the difficulty of explaining AI-driven decisions to affected stakeholders are not abstract concerns. A single high-profile failure, whether a biased hiring algorithm, a flawed credit decision, or a misleading customer-facing output, can cause significant and lasting damage to brand reputation and customer trust. Fairness, transparency, and accountability need to be built into AI systems from the outset, not added as an afterthought when something goes wrong.
Cybersecurity in the AI era demands a significant increase in sophistication. AI is expanding the attack surface and equipping adversaries with more capable tools. At the same time, AI offers powerful new defensive capabilities, from anomaly detection to automated threat response. A less visible but equally real risk is strategic overcommitment to platforms that are not yet mature. The large unifying AI platform plays are proving slower to deliver than many organizations anticipated. CIOs who have built significant dependency on a single vendor or architecture may find themselves exposed when that platform evolves more slowly than promised or takes a different direction altogether. Maintaining architectural optionality and ensuring that individual components can be replaced without unraveling the whole is both a technical and a risk management imperative.
Perhaps the most underestimated risk remains managing expectations within the business. AI has arrived with extraordinary levels of enthusiasm, and that enthusiasm is not always matched by a realistic understanding of what it can deliver, how quickly, and at what cost. Many pilots will fail. Many that succeed technically will struggle to demonstrate business value at scale. The CIO must be an honest voice in executive conversations, willing to communicate setbacks transparently and ensure the organization learns from what does not work as rigorously as it celebrates what does. Trust, once lost through over-promised and under-delivered AI initiatives, is difficult to rebuild.
Every AI ambition in your organization will eventually collide with the same reality: your data estate is probably not as ready as you think.
Make data your first investment, not your second thought. Every AI ambition in your organization will eventually collide with the same reality: your data estate is probably not as ready as you think. Before scaling any AI initiative, conduct an honest audit. Treat data readiness as the entry ticket to any serious AI program, not a parallel workstream. And set realistic expectations internally: the goal is not perfect data; it is building systems and processes that perform well under real-world conditions.
Stop piloting. Start governing. Most large organizations have too many AI experiments and too little discipline around them. To ultimately move faster with AI, map the full portfolio, apply consistent criteria, and kill what is not working. This requires having defined exit criteria before a pilot launch, not after it has consumed 12 months of goodwill and budget.
Architect for replaceability, not just functionality. Models will improve, vendors will pivot, and the AI landscape will continue to shift in ways none of us can fully anticipate. Build your AI architecture so that individual components can be swapped out without unraveling everything else. Modularity is not a technical preference. In the current environment, it is the only responsible way to build.
Own the governance of your machine workforce. AI agents are already working inside your organization, making decisions, executing processes, and in some cases talking to your customers. Most IT functions are not yet governing this properly. Treat AI agents as a new category of worker: define onboarding standards, access controls, audit trails, and legal accountability from the outset. The organizations that get ahead of this now will avoid the painful retrofitting that comes from ignoring it.
Position yourself as the organization’s AI conscience, not its AI owner. The CIO who tries to own the AI transformation agenda will become a bottleneck and, likely, a scapegoat. Your role is to enable, govern, advise, and challenge. Push accountability into the business functions where value is actually created. Coach your CEO and leadership team to own their AI outcomes. And be the voice in the room that tempers unrealistic expectations, calls out governance gaps, and keeps the organization honest about what is working and what is not. That kind of credibility, built over time, is worth considerably more than any individual AI initiative you could deliver.
The Chief Information Officer is becoming the Chief Intelligence Officer.
The Chief Information Officer is becoming the Chief Intelligence Officer. More than a simple rebranding, this reflects a substantive shift in what the role is actually for and what it will be judged on in the years ahead.
Now is the time to rethink the purpose of the IT function from the ground up, redefine the nature of CIO leadership, and fundamentally reposition the relationship between technology and the rest of the organization. The IT function will not fade into the background as AI matures and spreads. What will diminish is the routine, the commodity, and the work that required large teams precisely because it had not yet been automated. What will grow is everything that requires judgment, governance, architectural thinking, and understanding of how an organization creates value. The CIO who can govern data as a strategic asset, design architectures that remain flexible in the face of continuous and often unpredictable disruption, manage a workforce that increasingly blends human ingenuity with machine capability, and do all of this while maintaining the trust and confidence of the organization will be among the most consequential leaders in the enterprise. Decisions being made right now, about how to adopt AI responsibly, how to govern it rigorously, and how to build lasting organizational capability around it, will shape competitive positions and institutional resilience for years to come.
The Austrian crystal company achieved over 5% contribution to digital commerce with less than 1% of digital budget by emphasizing a value-first approach, its Chief Digital and Information Officer told IMD’s Luxury...
Hugo Travers built a 20M-strong news audience by prioritizing accessibility, trust, and transparency, offering leadership lessons for the digital age....
To unlock true competitive advantage, leaders must learn to integrate the refined intelligence of American AI models with the coordinating power of China. Howard Yu explains why most Western companies are on...
Audio available
AI excels at complex tasks, yet its effectiveness depends on strictly controlled boundaries...
Audio availableExplore first person business intelligence from top minds curated for a global executive audience