The paper’s practical point: AI transformation is not just a technology adoption problem. It is a leadership-choice problem.

Source note: Mika Ruokonen and Paavo Ritala. “Leading AI transformation: three approaches for CEOs.” Strategy & Leadership, published online February 9, 2026. DOI: https://doi.org/10.1108/SL-12-2025-0433. Emerald landing page: https://www.emerald.com/sl/article/doi/10.1108/SL-12-2025-0433/1339776

Why This Paper Matters

Most AI transformation advice quietly assumes that the leader’s job is to approve the strategy, fund the roadmap, and let technical people execute.

That is incomplete.

The hard question is not whether a company should use AI. The hard question is what kind of leadership system the company needs around AI. Some CEOs should create direction and room for specialists. Some need to become hands-on learners because the strategic choices are too technical to delegate cleanly. Some are operating in markets where AI is not an efficiency program, but a business-model rupture.

This paper is useful because it names those differences.

Ruokonen and Ritala study how CEOs and managing directors lead AI transformation across 31 Finnish organizations. They identify three broad approaches: Generalist, Expert, and Disruptive. The categories are simple, but the implication is sharp: a CEO’s personal orientation to AI shapes the operating model of the transformation.

That matters for boards, investors, operators, and product leaders. If the CEO treats AI as a delegated transformation program, the company will make different bets than if the CEO treats AI as a personal strategic agenda or an existential market shift. None of those positions is automatically right. Each creates a different set of strengths, blind spots, metrics, and failure modes.

The Idea in Plain English

The paper is basically a typology of CEO posture.

The Generalist CEO sees AI as important, but not as the center of their own job. They set direction, provide resources, remove obstacles, and trust specialists to lead much of the work.

The Expert CEO gets close to the technology. They invest personal time in understanding AI, lead by example, embed AI into management routines, and actively shape scope and pace.

The Disruptive CEO treats AI as a make-or-break shift. They use it to challenge the existing business model, reposition the company, and place bigger bets on new markets or operating models.

These are not personality types in the pop-business sense. They are leadership operating modes. A CEO can blend them, and the right mode can change as the market changes. The useful move is asking which mode the company is actually running, whether that mode fits the industry, and what failure pattern it creates.

What the Researchers Tested

The study uses 31 qualitative interviews with CEOs and managing directors from large organizations and ambitious growth firms in Finland. The sample spans construction, manufacturing, energy, finance, legal services, software, media, advertising, fashion, healthcare, pharmaceuticals, consulting, recruitment, public sector organizations, and other industries.

The paper is not trying to produce a universal quantitative model. It uses inductive thematic analysis to compare how senior leaders describe their role in AI transformation.

The public research summaries and the author deck emphasize several variables: CEO digital or AI expertise, market volatility, company size, sector, and tenure in the CEO role. The authors also collected secondary data to understand the pace and context of each firm’s AI transformation.

That design fits the question. AI transformation is still too early and too context-dependent for a tidy maturity model to explain everything. The more useful thing is to understand the recurring leadership patterns and the risks that come with them.

What They Found

There are three recognizable CEO approaches

The Generalist approach is traditional leadership adapted for AI.

These CEOs recognize AI’s potential, but they do not make AI their dominant personal priority. Their job is to create direction, allocate resources, define roles, choose partners and technologies, and make space for specialist-led execution. They are most useful when they align AI work with the core business instead of trying to become the company’s chief AI expert overnight.

The Expert approach is hands-on.

These CEOs have meaningful digital or AI expertise, or they deliberately build it. They lead by example, keep AI on the executive agenda, engage with technical alternatives, and push the organization to build capability. Their advantage is that they can make better strategic judgments because they understand the technology well enough to ask sharper questions.

The Disruptive approach is reinvention-oriented.

These CEOs treat AI as a strategic rupture. The goal is not only productivity or automation. It is to create new offerings, redesign the business model, challenge industry assumptions, and reposition the company. This approach is more aggressive and can be appropriate when the market is volatile or when AI changes the basis of competition.

The CEO’s AI literacy changes the possible strategy

One of the paper’s clearest claims is that personal commitment, enthusiasm, and expertise matter.

That does not mean every CEO must become an engineer. It means AI literacy has become a strategic competence. A leader who does not understand AI well enough to judge its timing, limits, and organizational consequences is likely to either overdelegate or overhype.

The R&D Today summary reports that market volatility, combined with the CEO’s own IT and AI knowledge, appears to influence the chosen approach. In stable markets, CEOs with less AI knowledge tended to act more like Generalists. In volatile markets, CEOs with stronger AI knowledge were more likely to pursue more Expert or Disruptive moves. The authors also stress that this relationship is not mechanical. There are exceptions, and real leadership behavior is blended.

Sector alone does not determine the leadership style

The tempting view is that software companies need Expert or Disruptive CEOs, while regulated or industrial sectors need Generalists.

The study is more nuanced.

All three approaches appeared across multiple industry segments. The author deck says company size showed little observable pattern in the qualitative data, and CEO tenure appeared to have limited relevance. Sector matters, but it does not dictate the answer.

That is important because many companies hide behind their category. A legal services firm may assume AI transformation should be slow and delegated. A software firm may assume it should be disruptive by default. The paper suggests the better question is whether the leadership approach fits the company’s volatility, readiness, capabilities, stakeholder pressures, and strategic intent.

Each approach has its own failure mode

The Generalist approach can fail through distance. If the CEO delegates too much, AI projects may become fragmented experiments owned by specialists but disconnected from company strategy. The company may optimize local use cases while missing a larger shift.

The Expert approach can fail through overconcentration. A CEO who is deeply committed to AI can drain attention and resources from other strategic priorities. They can also demoralize other leaders if AI becomes a personal crusade rather than an organizational capability.

The Disruptive approach can fail through overreach. Bold AI bets can become hard-to-reverse commitments. Internal and external stakeholders can grow tired of inflated promises. Traditional teams can feel displaced by the new AI agenda. The company may pursue market repositioning before it has the capabilities, trust, or operating discipline to absorb the change.

This is the paper’s most practical contribution. It does not say, “be more visionary.” It says: know which leadership posture you are taking, then manage the failure pattern that posture creates.

Why It Happens

AI transformation compresses technology, strategy, operations, and culture into the same decision space.

Older digital transformations could often be framed as systems projects: move to the cloud, digitize the customer journey, automate a process, modernize the data stack. AI reaches deeper into judgment, work design, knowledge flows, measurement, and the boundary between human and machine work.

That changes the CEO role.

If AI can reshape recurring routines, change how managers use data, alter employee autonomy, free up work time, and open new business models, then it cannot be treated only as an IT initiative. The leader has to decide what AI is for.

Is it a productivity layer? A capability-building agenda? A new basis of competition? A reason to redesign leadership practices? A threat that requires repositioning? A tool that should stay subordinate to the existing strategy?

Different CEOs answer those questions differently. Their answers become the operating model.

What This Means for Builders

Builders selling AI into companies should pay close attention to the CEO posture.

A Generalist buyer needs strategic alignment, governance, credible ROI, clear roles, and a path for specialists to execute without losing connection to the core business. The product has to help the organization coordinate AI work, not just demonstrate model capability.

An Expert buyer will ask better technical and operating questions. They may care about architecture, capability transfer, integration, and speed. They will be less impressed by generic demos and more interested in whether the product changes the organization’s ability to build and manage AI-enabled workflows.

A Disruptive buyer wants leverage against the business model. They are looking for repositioning, new revenue, new market entry, faster experimentation, and a story about why the company can move before competitors do. They will also need sharper risk controls because their appetite can run ahead of organizational readiness.

The product implication is straightforward: the same AI product may need three different sales and implementation motions. Selling one generic transformation narrative into all three leadership modes will miss the actual buying logic.

What This Means for Buyers and Operators

Buyers should diagnose their leadership mode before launching another AI program.

If the CEO is operating as a Generalist, the company should make sure AI initiatives have clear ownership, strategic fit, and enough executive attention to avoid becoming disconnected experiments.

If the CEO is operating as an Expert, the company should make sure personal conviction becomes organizational capability. The goal is not to make every decision flow through the CEO’s AI agenda. It is to distribute judgment, build management practices, and keep non-AI priorities visible.

If the CEO is operating as a Disruptor, the company should define risk tolerance early. That means deciding which bets are reversible, which are existential, which traditional teams need to be protected or retrained, and how the organization will avoid AI fatigue.

The paper also suggests that CEOs need new metrics. Generalist programs may lean on ROI, profitability, FTE savings, and revenue. Expert programs may track transformed roles, engagement, capability development, and time-to-market. Disruptive programs may need innovation success rate, AI-enabled revenue, and industry-positioning measures.

One metric set will not fit every leadership mode.

What to Watch Next

The field should watch whether AI leadership moves from enthusiasm to operating discipline.

The next serious question is not which CEOs are “pro-AI.” It is whether they can match their leadership posture to market volatility, organizational readiness, and strategic intent. A cautious Generalist can be right in one context and dangerously slow in another. An Expert can create real capability or become a bottleneck. A Disruptor can create a new market position or burn trust before the organization is ready.

Researchers should also watch blended leadership modes. The author deck notes that real leaders often span more than one category. That is likely to become more common as AI moves from experimentation to daily management practice.

Builders and investors should watch the handoff from CEO posture to execution system. It is easy for a CEO to declare an AI transformation. It is harder to redesign routines, incentives, measures, governance, employee trust, and customer promises around the chosen approach.

Limitations and Caveats

This is qualitative research based on 31 interviews, not a large causal study. The sample is centered on Finland, so the findings should be read as a useful typology rather than a universal law of AI leadership.

The categories are also porous. CEOs can combine Generalist, Expert, and Disruptive behaviors, and the right blend may change over time.

The public materials around the paper are unusually helpful, but the Emerald article itself was not directly accessible through the local CLI during this workflow because the page returned a Cloudflare challenge. This explainer relies on Crossref DOI metadata and public author, university, and practitioner summaries rather than a direct full-text extraction from Emerald.

That caveat does not weaken the core takeaway. The available sources consistently describe the same structure: three approaches, 31 interviews, cross-industry variation, and the importance of matching CEO leadership style to AI maturity and strategic context.

Source

Ruokonen, M., & Ritala, P. (2026). Leading AI transformation: three approaches for CEOs. Strategy & Leadership, 1-17. Available at: https://doi.org/10.1108/SL-12-2025-0433