Before AI spend becomes operational risk, leaders need to know who is genuinely accountable for the outcome.

​

​

The question to ask before approving AI​

 

AI projects often have a budget owner, a delivery team and a technical lead. That is not the same as having one named business owner who carries the outcome.

​

If ownership is unclear when a project is approved, it usually becomes even less clear when costs rise, adoption falters, delivery problems appear or trust is damaged.

​

This check is designed for use before approving an AI pilot, platform, workflow change or automation project where accountability is still vague.

​

​

The ownership question

 

​If this project underperforms, disrupts work, costs more than expected or damages trust, which leader owns the consequence?

​

A sound AI decision names the person who owns the business outcome, not simply the person managing the project.

 

Three ownership decisions to make

 

1. ​Name the outcome owner. 

Who is accountable for the business result this AI project is meant to create?

The outcome should be specific enough to explain in plain English, and its owner should have enough authority to change the plan rather than simply report on it.

​

2. Name the adoption owner.

Who is responsible for whether the people affected by the change can actually use it, trust it and work with it?

Technical delivery does not guarantee adoption. Someone must own what happens when the technology meets the organisation.

3. Name the pause authority

Who has permission to reduce the scope, redesign the approach or stop the project when the evidence is not holding?

A pause authority must be able to act. It cannot be someone who is expected only to raise concerns while the project continues unchanged.

​

 

What good ownership looks like

​

Good ownership means:

​

• The intended outcome is specific enough to explain in plain English.

• One named leader is accountable for the business result.

• The owner has enough authority to change the plan.

• Responsibility for adoption and its effect on people is clear.

• The leadership team agrees what evidence would demonstrate that the project is working.

• Someone has explicit authority to reduce the scope, redesign or stop the project.

• The decision can be defended financially, operationally and reputationally.

​

 

The five question leadership check

​

Use these questions with your leadership team before approving the project. If an answer is vague, leave it blank and treat that as useful evidence.

​

Business outcome
What measurable result is this AI project expected to improve?

​

Named owner
Who owns the result if the project underperforms?

​

People and adoption impact
Who will use, check or trust the system, and who will be affected by the change?

​

Pause authority
Who can reduce the scope, redesign the approach or stop the project?

​

Evidence to review
What will you measure, and when will you review it?

 

​

If the answer is unclear

​

If you cannot name the owner, the decision is not ready.

​

Do not treat unclear accountability as something to resolve after implementation. Treat it as evidence that the organisation needs to take a Human Pause before approval.

​

​