What is the role of leadership in AI transformation?

Leadership in an AI transformation is responsible for its adaptive challenges, the parts of the change with no known answer. Leaders frame the questions, set the pace and protect the people doing the work while they learn. What that involves depends on whether the organization wants faster work, new offers or a different business model.

What is the difference between a technical and an adaptive challenge?

A technical challenge has an answer that someone already knows, so bringing in the right expertise solves it, even when the work is difficult. With an adaptive challenge, nobody has the answer yet. The people involved have to learn, experiment and change how they work to make progress, and no expert can do that for them. Most AI initiatives contain both.

How should managers support AI adoption in their teams?

Managers support adoption by working with the tools themselves first, naming what people will lose as well as what they will gain, making it acceptable to be slower while people learn, and changing what gets measured and rewarded. These matter most on projects that aim to make work faster, because the work being automated is often work people were proud of doing well.

Does the leader’s role change with the type of AI project?

Yes, it changes with the kind of value the project is after. On projects that make existing work faster, most of the challenges are technical, and the leader deals with what people lose and changes what gets measured. Projects that create new offers are roughly half adaptive and need the leader to define good experiments and stop the ones that fail. Projects that challenge the business model are mostly adaptive, and there the leader’s main job is to hold an unresolved question open while people explore it.

Method

The leader’s role in AI transformation

A leader’s role in AI transformation is to lead the parts of the change that have no ready answer, where people have to learn new ways of working and give something up. What that involves depends on the kind of value an AI initiative is after: faster work, new offers for customers, or a different business model.

Two ideas to start with

The leader’s role rests on two ideas that Think Big Leaders uses throughout its work on AI: the difference between technical and adaptive challenges, and the AI value framework, which groups AI initiatives by the value they are meant to create.

Technical and adaptive challenges

A technical challenge is one where somebody already knows the answer. The knowledge exists, inside the organization or outside it, and it can be hired, bought or taught. An adaptive challenge has no known answer yet. Progress depends on the people involved changing what they believe, what they value or how they work, so it cannot be handed to an expert. Technical does not mean easy, and most real changes contain both kinds.

Area	A technical challenge	An adaptive challenge
Medicine	A heart bypass. It takes years of training and the stakes are high, but the patient does not have to change.	Getting the same patient to change how they eat, sleep and handle stress once they are home.
Machinery	Tracing an intermittent fault in a complex machine. Difficult expert work, with one right answer to find.	Deciding to stop maintaining the machine, because the business no longer needs what it makes.
Sport	Rebuilding a player’s technique. The coach knows how and drills the player in it.	A team that wins but will not pass to each other, or a senior player whose role has to shrink.
AI	Rolling out a tool and training people to use it.	Getting people to change how they work, what they trust and what they stop doing by hand.

The distinction comes from Adaptive Leadership, developed by Ronald Heifetz and his colleagues at the Harvard Kennedy School. Think Big Leaders built its own framework for leading AI transformation on it, and uses that framework in its leadership workshop.

Three paths to value from AI

Our AI value framework puts every AI initiative on one of three paths. The paths differ in who owns the work, how long the first result takes, and how much of the change is adaptive.

1
Boosting productivity
Perform tasks and workflows faster, at larger scale, higher quality, and lower cost.
2
Creating new value
Build innovative solutions to accelerate time to value for your stakeholders.
3
Driving disruption
Challenge value chains, business models and modes of operation.

The share of adaptive challenges grows from the first path to the third: productivity work is mostly technical, and disruption is mostly adaptive.

What AI, the team and the leader do on each path

Path	What AI can do	What the team does	What the leader does
Boosting productivity	Drafting, summarizing, searching, reconciling, translating, editing, triaging, making data accessible and visualizing it.	Codifies best practice, validates the output and is accountable for it, and gives up a craft people were proud of.	Uses the tools in their own work first. Names the loss as well as the gains. Makes it safe to be slower before faster. Changes what gets measured and rewarded.
Creating new value	Exploring options, prototyping, testing, building and operating at a cost that used to be prohibitive.	Decides what is worth creating, builds it together with AI, matches AI capabilities to real needs, and keeps the focus on customers and stakeholders.	Defines what a good experiment and a good failure look like. Protects the work from short-term pressure. Stops the bets that are not working.
Driving disruption	Gathering evidence, research, scenarios and comparisons, running experiments, and making new operating models possible.	Challenges current business models and value chains, tells leadership what is happening on the ground, and monitors, gives feedback and iterates.	Discusses the disruption openly. Frames the questions that need exploring. Defines what does not change and what to let go of. Holds the tension without resolving it too soon.

The leader’s three main moves

A leader makes the same three moves on every path, often within one conversation and more than once.

Diagnose the work: Split the challenge into its technical and adaptive parts, then work out who has to learn something and what they will have to give up.
Regulate the heat: Keep enough pressure on people for them to move, but not so much that they freeze. Raise it by naming what is at stake, and lower it by breaking the work into smaller steps and setting a pace.
Give the work back: Frame the challenge and let the people who have to change work out the answer. If your instruction to the team still contains the solution, you have delegated the task and kept the problem.

What the moves involve changes with the path, as the leader column in the table shows. There is more on each move on our leading AI transformation page.

Using the table with a leadership team

1
Place each initiative on a path
List the AI initiatives the team is leading and put each one on the path it mainly serves.
2
Compare with the leader column
For each initiative, compare what the leadership team is doing now with the leader column for that path.
3
Pick the missing move
Choose one leader behaviour that is missing, and agree who will start doing it and by when.
4
Look again next quarter
Repeat this each quarter, because initiatives change path as they mature.

How we use the table in engagements

Leadership workshops: In the Leading Transformation in the Age of AI workshop, leadership teams and managers place their own initiatives on the paths and practice the three moves on real challenges.
Discovery work: The AI value framework decides which initiatives to pursue. This table follows it, to agree what the leader does on each one.
Adoption programs: The team column shapes the training and champions work, and the leader column shapes what managers are asked to do differently.

Common questions

What is the role of leadership in AI transformation?: Leadership in an AI transformation is responsible for its adaptive challenges, the parts of the change with no known answer. Leaders frame the questions, set the pace and protect the people doing the work while they learn. What that involves depends on whether the organization wants faster work, new offers or a different business model.
What is the difference between a technical and an adaptive challenge?: A technical challenge has an answer that someone already knows, so bringing in the right expertise solves it, even when the work is difficult. With an adaptive challenge, nobody has the answer yet. The people involved have to learn, experiment and change how they work to make progress, and no expert can do that for them. Most AI initiatives contain both.
How should managers support AI adoption in their teams?: Managers support adoption by working with the tools themselves first, naming what people will lose as well as what they will gain, making it acceptable to be slower while people learn, and changing what gets measured and rewarded. These matter most on projects that aim to make work faster, because the work being automated is often work people were proud of doing well.
Does the leader’s role change with the type of AI project?: Yes, it changes with the kind of value the project is after. On projects that make existing work faster, most of the challenges are technical, and the leader deals with what people lose and changes what gets measured. Projects that create new offers are roughly half adaptive and need the leader to define good experiments and stop the ones that fail. Projects that challenge the business model are mostly adaptive, and there the leader’s main job is to hold an unresolved question open while people explore it.

Further reading: Amir Elion writes about leadership and AI transformation at amirelion.com.

Run this with your leadership team30 min or send a message