← Back to all articles

How to Keep AI Transformation Moving When Leadership Above You Is Shifting

By Vance Sterling·9 min read·April 1, 2026

Nobody writes about this part. Every AI transformation article assumes you have stable executive sponsorship, a clear mandate, and a leadership chain that agrees on direction. In 20 years of enterprise IT at major banks, I have had that luxury exactly zero times. The real skill is keeping a transformation alive when the people above you are fighting, getting replaced, or sending contradictory signals every other week.

The Instability Tax on AI Transformation

Most AI projects die from the top down, not the bottom up. A 2024 BCG study found that 74% of stalled enterprise AI initiatives cited 'shifting executive priorities' as a primary blocker. Not technical debt. Not talent gaps. Leadership instability.

Here is what that looks like in practice. You are six weeks into a command center transformation. The CTO has signed off on an AI-first vision. Your team is aligned and building. Then your boss fails an audit. His boss starts circling. Suddenly the person who approved your budget is worried about survival, not your project roadmap.

I call this the instability tax. It is the 20-40% productivity drain that hits every initiative when leaders above you are distracted, threatened, or replaced. Your team does not stop working, but they start hedging. Meetings get longer. Decisions get deferred. People start updating their resumes instead of updating their deliverables.

The instability tax is invisible on project dashboards. No one logs 'spent 3 hours worrying about reorg' as a status update. But it compounds fast. Two weeks of instability costs you a month of momentum.

The Three Things You Actually Control

When leadership above you is volatile, your instinct is to wait. Wait for clarity. Wait for the dust to settle. Wait for the new boss to get oriented. That waiting kills more transformations than any technology failure ever has.

Instead, focus on the three things you always control regardless of who sits above you: your team's daily output, your documentation trail, and your stakeholder map.

Daily output means your team ships something visible every week. Not a slide deck. A working prototype, a completed integration, a measurable result. When leadership changes, the first question the new person asks is 'what has this team actually delivered?' If your answer is 'we are still in planning,' you are dead. If your answer is 'here are 11 completed automations saving 140 hours per month,' you survive every reorg.

Documentation trail means every decision, every approval, every scope change is written down with dates and names. I learned this the hard way at a Top 5 bank when a new CIO arrived and asked why we were spending $2.3M on an automation platform. The previous CIO had approved it verbally in a hallway conversation. No email. No memo. We nearly lost the entire program. Now I document everything as if a hostile auditor will read it tomorrow, because eventually one will.

Stakeholder map means you know exactly who benefits from your project at every level. Not just your boss. Your boss's peers. The business unit leaders whose teams get time back. The compliance officers whose audit findings get resolved. When your direct sponsor disappears, these lateral relationships become your lifeline.

The 72-Hour Leadership Transition Framework

I have been through nine leadership transitions during active transformation programs. Some were planned retirements. Some were Friday afternoon terminations. Each time, I had about 72 hours before the new leader started forming opinions about what to keep and what to kill.

Here is the framework I use in those 72 hours. Day one: send the new leader a single page. Not a 40-slide deck. One page with three numbers. What the project costs, what it has delivered so far, and what it will deliver in the next 90 days. Attach a one-paragraph summary. Nothing else. New leaders are drowning in information. The person who gives them clarity wins.

Day two: brief your team. Tell them what you know, what you do not know, and what will not change regardless of who is in charge. Be specific. 'Our sprint goals for the next two weeks are unchanged. Our Q2 milestones are unchanged. Budget conversations may shift but I will handle those.' People can handle uncertainty. They cannot handle silence.

Day three: activate your stakeholder map. Reach out to every lateral ally and ask them one question: 'If the new leader asks you about our project, what would you say?' This does two things. It reminds them your project exists. And it gives you early warning if anyone is planning to throw your initiative under the bus to protect their own.

This framework has saved three different programs I have led from getting axed during transitions. The one time I did not follow it, at a bank in 2019, we lost eight months of work because the new SVP had no context and defaulted to cutting everything the previous leader had approved.

Building Change That Survives Its Sponsors

The best change management strategy assumes your sponsor will leave. Not because you are pessimistic, but because the average tenure of a bank CTO is 4.3 years and the average AI transformation takes 3-5 years. The math does not work in your favor.

So you build the transformation to be sponsor-independent. That means three structural decisions from day one.

First, tie every AI initiative to a business metric that matters to the business unit, not to IT. 'We reduced false positive alerts by 62%' survives a leadership change. 'We implemented the CTO's AI vision' does not. When the CTO leaves, so does their vision. But a 62% reduction in false positives is just a fact.

Second, distribute ownership across the team. If you are the only person who can explain the transformation strategy, you are a single point of failure. I run my current command center transformation with six workstream leads who each own a piece of the vision. If I get moved tomorrow, the work continues. That is not selfless. That is engineering for reality.

Third, create visible wins on a 30-day cycle. Not quarterly. Monthly. Leadership transitions happen on random Tuesdays, not at the end of fiscal quarters. If your next deliverable is four months away, you are exposed for four months. If your next deliverable is three weeks away, you always have something fresh to show whoever walks through the door.

I am running this playbook right now with a major transformation. The honest truth is I do not know what happens next with my leadership chain. But my team is shipping. Our results are documented. Our stakeholders know what we deliver. Whatever happens above me, the work speaks.

Free Resource

Want the Complete AI Leadership Playbook?

The complete AI leadership framework — strategy, governance, and implementation in one book.

Get the Book on Kindle

Actionable Takeaway

This week, write a one-page brief for your most important AI initiative. Three numbers: total cost, value delivered to date, and projected value for the next 90 days. Keep it updated monthly. When leadership shifts, and it will, you will have clarity ready while everyone else scrambles to build a deck.

This article covers a core framework from The Executive's AI Playbook. The complete playbook includes printable scorecards, additional real-world examples, and full implementation checklists.

Get the complete framework →

Get the AI-First Leader on Kindle

The complete AI leadership framework — strategy, governance, and implementation in one book.

Not ready to buy?

Start free: 5 AI Questions Every Executive Must Answer Before Investing →

Free PDF guide. No spam.