In March 2026, Coca-Cola CEO James Quincey told CNBC that AI had significantly influenced his decision to step down. The company needed, in his words, "someone with the energy to pursue a completely new transformation of the enterprise." A few months earlier, Walmart's Doug McMillon had stepped aside for essentially the same reason. He could start the next set of AI transformations, he said, but he could not finish the job. Walmart needed someone faster.
Neither man was pushed out. Quincey had added more than ten new billion-dollar brands to Coca-Cola during his tenure. McMillon had led Walmart through over a decade of sustained growth. Both concluded, independently, that the AI era demanded a kind of leadership they could not provide.
What Quincey and McMillon recognized is something most leadership teams have not yet confronted. The AI era does not just demand new technology or new strategy. It demands different leadership, with specific competencies and behavioral patterns that differ in important ways from what prior eras required.
The harder version of this challenge belongs to organizations. A CEO stepping aside and passing the baton is, in a narrow sense, a clean solution. An organization does not have that option. It has to develop the leadership it needs, systematically and at scale, or it will fail with the leadership it has.
The 90-day framework below is designed to start that work.
What AI-Era Leadership Actually Requires
Before assessing or developing leadership against AI-era criteria, it is worth being precise about what those criteria are.
AI-era leadership requires a different relationship with uncertainty. Decisions get made with incomplete information, against a competitive landscape that shifts faster than annual planning cycles, and often before best practices have been established. Leaders who need to be confident before committing will be consistently late.
It requires genuine technical fluency. Not the ability to write code, but the ability to reason accurately about what AI systems can do, what they fail at, what they cost, and what risks they introduce. Leaders who rely entirely on technical teams to translate for them lose the ability to make informed strategic decisions.
It requires comfort delegating to non-human systems. An executive who insists on reviewing every AI-generated output before it is used, or who will not allow consequential decisions to be influenced by AI reasoning, is not a meaningful bottleneck. They are a structural impediment to the organization's ability to operate at AI speed.
And it requires a bias toward experimentation over optimization. The organizations gaining competitive ground on AI are not the ones that designed the perfect implementation plan. They are the ones that launched, failed fast, and iterated. Leadership cultures that treat failure as a career risk cannot produce this behavior at scale.
Days 1–30: Assess
The goal of this phase is an honest picture of where your leadership team stands, not where they think they stand, and not where they told the board they stand.
Assess AI fluency across the senior team. Run a structured assessment of every member of the senior leadership team against a defined fluency rubric covering foundational understanding of how AI systems work, awareness of their failure modes, command of the cost and risk implications, and ability to connect AI capability to business strategy. The output is a fact-based gap map, not a ranking.
Diagnose mindset gaps. Fluency is a knowledge problem. Mindset is a behavioral one. Assess each leader against specific behavioral markers: tolerance for ambiguity, willingness to shut down their own initiatives when data suggests they are not working, comfort delegating to non-human systems, and genuine bias toward experimentation. The goal is not to grade leaders; it is to identify which behavioral patterns will accelerate transformation and which will block it.
Map decision-making patterns. Examine the last ten significant decisions your leadership team has made. How long did each take? How much information was gathered before committing? How often were decisions revisited? How many were reversed? The pattern that emerges tells you whether your leadership team is structurally capable of operating at the pace the AI era demands.
Stress-test the CEO. The tone is set at the top. If the CEO is not personally fluent in AI, not personally using AI tools, and not personally comfortable with ambiguity and failure, the rest of the organization will not take transformation seriously. The CEO's development plan must be the most rigorous of any member of the leadership team.
By the end of Day 30, you have a clear and evidence-based picture of your leadership team's AI fluency, behavioral readiness, and the specific individual and collective gaps that the next phase needs to address.
Days 31–60: Develop
This phase builds the capabilities and behaviors identified as missing, not through generic leadership training, but through deliberate, role-specific development tied directly to the decisions each leader is responsible for.
Build individual development plans. Every member of the senior leadership team needs a written development plan tied to the specific gaps the assessment revealed. The plan should specify target capabilities, the activities that will build them, and the measurable outcomes that demonstrate progress. Generic leadership curricula do not close AI-specific gaps. The plan must be specific to the leader and to the decisions their role requires them to make.
Put AI to work in their actual jobs. Fluency does not come from reading about AI. It comes from using it. Every senior leader should be actively using AI tools in their daily work by Day 45: drafting with them, analyzing data with them, stress-testing strategy against them. Usage is not optional.
Run decision simulations. Design AI-era decision scenarios specific to your industry and strategic priorities, then run your leadership team through them. The scenarios should force confrontation with decisions the team is currently avoiding: when to let an AI system make a consequential call autonomously, how to handle workforce transitions, how to respond when a competitor deploys AI faster than you can respond. Judgment is built by exercising it under conditions that approximate the real thing.
Build peer learning structures. The fastest leadership learning happens in small groups of peers confronting similar challenges. Pair each senior leader with one or two peers, inside or outside the organization, who are working through comparable AI decisions. These groups should meet on a defined schedule and work through real-world cases.
Expose leaders to the frontier. Your leadership team must have structured, regular exposure to the state of the art: not to the state of the market, which always lags. That means direct engagement with AI labs, leading researchers, and organizations further along in deployment. Leaders who only see what their vendors are selling will consistently underestimate what the competition is doing.
Realign evaluation criteria. If the leadership evaluation framework has not changed in five years, the behavioral expectations have not actually changed regardless of what the strategy documents say. Tie a meaningful portion of leadership evaluation to AI-readiness indicators: experiments personally sponsored, fluency demonstrated in board-level discussions, talent developed in the direction the organization needs to move.
By the end of Day 60, every senior leader has a development plan in motion, is using AI tools directly, has been tested through decision simulations, and is being evaluated against criteria that reflect what the organization actually needs.
Days 61–90: Embed
This phase locks the changes into the operating fabric of the organization so that AI-ready leadership becomes a permanent characteristic rather than an initiative whose effects fade within a quarter.
Embed AI into the leadership operating cadence. Every senior leadership meeting should include a structured AI component: a decision being tested, a capability being reviewed, or a risk being assessed. This is not a standing agenda item to be dropped when time runs short. It is a permanent feature of how the leadership team operates.
Rewire succession planning. The leaders your organization needs in three years are not the same as those it needed three years ago. Revisit the succession bench against AI-era criteria. Who is building genuine AI fluency? Who has stalled? The answers reshape where development investment goes and who gets flagged for accelerated development versus managed transition.
Build the board's fluency. A leadership team moving faster than its board will eventually slow to the board's pace. Build a structured AI education program for board members. At minimum, the board should have one director with deep AI expertise, a recurring agenda item for AI strategy and risk, and a shared vocabulary that supports substantive oversight rather than surface-level review.
Institutionalize the feedback loop. By Day 90 you have real evidence. Which development interventions changed behavior? Which leaders moved? Which did not? Use the data. Double down on what is working and redesign what is not.
Make the hard personnel calls. By this point, you know which members of the leadership team will make the journey and which will not. Delaying those decisions is not caution. It is cost: in execution, in culture, and in the talent retention of the people who are moving and watching whether leadership will act.
What a 90-Day Plan Does Not Solve
The 90-day framework produces evidence and sets development in motion. It does not complete the transformation.
Some of what the assessment reveals will require more than development. Leadership teams confronting AI transformation frequently discover that the gaps are not only in fluency or mindset but in structure: roles that were designed for a different operating environment, decision rights that are misaligned with how the organization now needs to move, and incentive structures that still reward the behaviors the transformation is trying to change.
These are not development problems. They are design problems. And the leadership decisions required to address them are among the most consequential in the AI era.
The 90-day plan creates the foundation to make those decisions with evidence rather than assumption. Organizations that use it well will know, by Day 90, where they have the leadership to go and where they need to build it.
Frequently Asked Questions
What does AI-ready leadership mean in 2026?AI-ready leadership describes a specific set of competencies and behavioral patterns: genuine technical fluency with AI systems and their failure modes, comfort delegating to non-human systems, tolerance for ambiguity and experimentation, and the ability to connect AI capability to strategic decisions. It is distinct from general technology literacy or change management experience, and organizations should assess against these specific criteria rather than using proxy measures.
How do Quincey and McMillon's departures inform leadership development strategy?Both executives made honest assessments of their own capacity to complete AI transformation and acted on those assessments. The organizational implication is that leadership development strategy must include equally honest assessment of the entire senior team, not just the CEO, and must address both fluency gaps and behavioral gaps. The leaders who succeed in AI transformation are not necessarily the most technically sophisticated. They are the ones willing to change how they work.
How long does it take to build a genuinely AI-ready leadership team?The 90-day framework creates foundational assessment, puts development in motion, and embeds the structural changes needed to sustain it. Genuine AI fluency and behavioral transformation at the senior leadership level takes 12 to 24 months of sustained effort. Organizations that treat the first 90 days as the finish line rather than the starting point consistently underperform on AI adoption.
What is the biggest mistake organizations make in AI leadership development?The most consistent mistake is skipping assessment and moving directly to development. Generic AI training programs delivered to leadership teams without a specific diagnosis of individual and collective gaps produce activity rather than capability change. The organizations that develop AI-ready leadership most efficiently start with an honest, evidence-based assessment and build development plans from that evidence rather than from curriculum catalogs.
How should succession planning change for the AI era?Succession planning in 2026 needs to evaluate candidates against AI-era criteria: demonstrated AI fluency, behavioral readiness for ambiguity and experimentation, track record of sponsoring AI initiatives, and ability to operate effectively in environments where non-human systems influence consequential decisions. Organizations using succession criteria designed for prior leadership eras are building benches that will be misaligned with the demands of the roles they are filling.