The Scouting motto—be prepared—has new meaning in the AI era.
Today, organizations are more likely to outperform competitors when their top executives are well prepared—when they are ready to meet the opportunities and challenges of the day and have what it takes for long-term success. In fact, organizations whose executives are well prepared gain three times the competitive boost they get from having a more mature AI deployment, according to new BCG research. (See the sidebar, “About Our Research.”)
About Our Research
Despite the critical role that well-prepared executives play, only 25% of employees feel that leaders adequately support their AI use. The deficit is one sign that too few executives have mastered the new leadership, technical, and responsible AI capabilities they need for the future. Even when they have, it has often occurred in a disjointed way.
Closing the gap is crucial because the stakes are unusually high. AI capability is compounding faster than most executives’ comfort with it, compressing formerly years-long decision cycles into mere months. Executives are being asked to commit real capital and to reorganize real teams around a technology that continues to change shape. Separate BCG research found that accelerating AI is a top-three priority for 65% of CEOs seeking to improve growth and productivity.
Our global survey found that those who are most ready to lead in today’s breakneck AI era are three dimensional: they assume connected roles as visionary, codeveloper, and steward, internalizing the capabilities and activities required for each dimension and alternating between them as the situation demands. By following in their footsteps, other executives can position themselves and their organizations to gain in performance and measurable impact.
AI Readiness Is Essential, But Most Execs Are Unprepared
Our survey found that executives are challenged by their organization’s AI transformations, which typically occur in three progressively more complex phases: Deploy, Reshape, and Invent. When leaders are challenged, AI maturity can suffer, as can the organization’s ability to scale the technology across multiple functions and use it to generate business value.
Given how much is required, it’s understandable that executives may not feel as prepared as they could be. In our survey, only 36% of leaders said that the senior leaders in their organization are prepared to meet today’s challenges.
But executives who are more prepared flex in and out of the visionary, codeveloper, and steward dimensions across AI transformations’ three phases, integrating leadership, technical, and responsible AI capabilities into each. When leaders have these capabilities, their organizations benefit. As previously stated, when an organization’s leaders are well prepared, their competitive outperformance is threefold compared with when their AI deployment is highly mature—66% v. 20%. When leaders are well prepared and have very mature AI practices, an organization’s competitive performance increases by an additional 80%. (See Exhibit 1.)
The Three Dimensions of AI-Ready Leaders
When executives effectively inhabit each of the three dimensions, clear gains in performance for their organizations follow.
THE VISIONARY
When operating in visionary mode, executives paint a clear picture of their organization’s AI vision and the goals needed to realize it. In this way, they help their teams adopt AI faster, and their organizations are likely to have more mature AI practices and be more adaptable. In many respects, this is in keeping with behavior change research described in Tomorrowmind that finds the ability to plan for the future to be a key differentiator of leaders whose teams thrive in uncertainty.
The visionary connects AI to what matters. Well-prepared, visionary executives incorporate AI into their overall business strategy. They are tech savvy enough to understand what AI can and cannot do and to anchor AI investments in measurable business outcomes. We found that mature AI organizations are up to three times more likely to have leaders who devise and communicate a clear, regularly refined AI vision. (See Exhibit 2.)
The visionary bets on optimism. Executives who are optimistic about what AI can do acknowledge its risks but lean into the opportunities it represents, allowing for a bigger, bolder AI vision. We found that executives who are strongly optimistic are 1.9 times more likely to lead organizations with mature AI practices.
The visionary leads differently at every phase. Highly effective executives change their leadership style as an organization’s AI maturity evolves. They may assume a more direct style during the Deploy phase, shift to enabling people and removing roadblocks as capability builds during the Reshape phase, and challenge assumptions or play devil’s advocate as the organization reaches the next horizon during the Invent phase. Our research found a statistically significant correlation between adaptability and AI maturity: 52% of leaders report that leaders at their organization can shift their management style depending on the situation.
THE CODEVELOPER
AI-ready executives acting in the codeveloper dimension tap into their technical know-how to model the new behaviors and work practices they want to see in their teams. They treat employees as codesigners, are open to experimentation, and use AI in their own work.
The codeveloper builds with, not for. Treating employees as codesigners while reshaping work itself means executives must create an environment where it feels safe to experiment and to address concerns about how AI could affect people’s work and futures. Treating employees as codesigners echoes a well-documented behavior pattern in which people place substantially higher value on the things they help create than on identical things handed to them fully formed.
According to How Change Really Works, a book by three BCG partners, three types of codesigning experiences give employees that kind of agency: decision making, influence, and representation. Well-prepared leaders make room for all three, so people feel like they’re part of the change. We found that organizations whose executives put people at the center of AI transformations are three times more likely to become AI mature (58% vs. 18%), unlocking value that can be scaled. (See Exhibit 3.)
The codeveloper uses AI in the work. Using AI builds intuition that leads to sound judgment. If leaders never use AI themselves, they cannot credibly assess where it helps or falls short or understand the guardrails it requires. Leaders who use AI on a day-to-day basis in their own work are more likely to lead AI-mature organizations: 45% compared with only 19% of those who never or rarely use AI.
The codeveloper makes room to experiment. A culture that supports experimentation promotes learning about and adopting AI in the most inventive ways. Executives who encourage AI experimentation throughout the organization build capability and buy-in. Our research found that leaders at AI-mature organizations are 75% more likely to experiment with the technology regularly and intentionally, to seek feedback, and to adapt: 40% vs. 23% at less AI-mature organizations.
THE STEWARD
When acting in the steward dimension, well-prepared executives concern themselves with AI’s boundaries. They think about guardrails that preserve fairness, privacy, quality, and safety and about the best way to shepherd their organization into the AI era.
The steward sets the guardrails. Executives recognize the importance of being clear about what AI can and cannot do and how it creates business value. While most executives underperform when it comes to creating clarity about AI, leaders who are effective in setting AI guardrails make sophisticated ethical and strategic decisions about where the technology should and should not operate, and why. Leaders who institute clear, human-centered guardrails are two to three times more likely to lead AI-mature organizations.
The steward makes risk-aware bets. Highly effective AI leaders use the technology to support core executive tasks such as making decisions, prioritizing, and thinking strategically. Executives who use AI to stress-test a major decision are not outsourcing judgment, they are deploying it more rigorously. They know a tool’s failure points, weigh its ability to make consequential decisions, and bring the output to the team to discuss in a way that builds confidence.
This mirrors existing research on how change really works—specifically on executive decision making—that shows how cognitive biases such as overconfidence can distort forecasts for major initiatives and how, as a result, executives can underestimate risk and overstate potential benefits. As AI expands leaders’ decision-making options, explicit guardrails and structured checks matter more, not less. We found that 38% of AI-mature organizations are led by executives who frequently use AI in their own decision making, compared with just 16% that are led by executives who rarely or never use it, a statistically significant relationship.
The steward never stops learning. Executives who spend the most time—more than 10%—teaching themselves how to use AI in their own work are 1.7 times more likely to lead AI-mature organizations.
And what leaders learn matters as much as how much they learn. The most effective learners continuously update their leadership approach, technical know-how, and understanding of responsible AI. (See Exhibit 4.)
How to Adopt the Practices of Well-Prepared Leaders
Executives who want to emulate the practices of well-prepared leaders can adopt their ways of thinking, leading, and learning.
To begin, assess the status of the organization’s AI transformation to understand which dimension to focus on first. Quantitative leadership surveys assessing the baselines of the visionary, codeveloper, and steward dimensions are a good place to start. Bringing in an external executive leadership coach can be useful, too; an outside perspective can often uncover blind spots that executives may not see in themselves. Whatever the starting point, pick a few practices to focus on initially, weave them into day-to-day work, and use language and repetition to create habits. Getting coaching support can accelerate this stage, providing accountability and repetition.
For the visionary dimension:
- Put value before volume. Anchor AI initiatives to measurable business outcomes before approving them. Make sure you are clearly imagining and then articulating how AI should serve the business, and that your team is aligning AI initiatives accordingly.
- Register but don’t avoid risk. Treat AI risks as concerns to be managed, not reasons to pause. Assign owners to investigate concerns and establish timelines for the work. Do not use caution as the default.
- Assess and adjust. Use a simple maturity framework to assess where a team is in an AI transformation and match your leadership style accordingly. Develop questions to ask for each phase. At the Deploy phase, it could be “What are we learning?” During the Reshape phase, it could be “What are we scaling?” In the Invent phase, it could be “What do we need to rethink?” When a team graduates to the next phase, make your own shift visible. Tell them: We are moving from exploring to building, and here is how my role will change.
For the codeveloper dimension:
- Use AI yourself. Commit to using AI to complete at least one meaningful work task a week. Talk publicly about how you use it to show people that you’re all in it together.
- Offer protected hours. Allocate dedicated time weekly that is protected from other priorities for teams to experiment with AI tools. Pair this offering with a simple debrief ritual to capture and share what worked.
- Maintain an open-door policy. Create space for employees to ask questions or voice concerns about AI. If questions come up that you don’t know how to answer, model vulnerability and a growth mindset and commit to finding a solution. Leaders who name and tame fear take away its power.
- Fail fast and share faster. Create a lightweight mechanism such as a communications channel or a monthly session to share experiment results, including failures. Learning, not just winning, is the asset.
For the steward dimension:
- Draw the line. Establish explicit AI decision rights and publish them to the team or organization you lead. Define where AI can act autonomously, where it can only advise, and where human judgment is non-negotiable.
- Bring the outside in. Hold regular sessions with external experts, including researchers, practitioners, ethicists, and vendors, to expose the leadership team to emerging AI developments. Regular learning can push people to think beyond the status quo and avoid putting a ceiling on ambition.
Highly effective AI leaders see leadership, technical prowess, and governance as parts of the whole. Being optimistic about the future means inspiring people with a positive vision but accounting for risk. Doing the work takes technical engagement and the credibility it builds. Establishing sufficient AI guardrails requires ethical judgment and courageous leadership. The executives who are winning today developed this fluency through integrated practice in the context of real work.
The authors wish to thank Tripp Twyman and Adriana Dimitriou for their contributions to this article.
