We need to make progress, but no one has the time or experience to lead it

Your organisation may already accept that AI deserves attention. The harder question is who leads it. With no internal AI capability and no senior person free to take it on, the work sits between roles. It comes up regularly, belongs to nobody and slips again whenever something more urgent lands.

That gap has consequences. Decisions get deferred, teams try things separately, and suppliers begin shaping the direction because nobody inside has written down what the business needs.

Why AI progress stalls without someone to lead it

A credible idea still needs someone to carry it. Somebody has to define the work, get the right people in a room, make the decisions that follow and keep it moving alongside everything else on their desk. Where that job has not been given to anyone, the initiative rarely fails outright. It just stops.

Signs you may recognise

• The senior team agrees AI is worth exploring. Nobody is accountable for the next step.
• The people best placed to lead it already have full diaries.
• No one internally has done this before, so decisions are deferred until someone feels more confident.
• Two or three teams are testing tools separately and nobody has the full picture.
• Supplier conversations start before the business has agreed what it actually needs.
• Privacy, security, governance and human oversight are treated as questions for later, with no named owner.
• There are no measures, so nobody can say whether anything has improved.

What follows is activity without decisions. Demonstrations get booked, articles circulate, a licence or two gets bought on somebody’s card. Months on, the organisation knows more about AI and no more about what it intends to do.

Ownership is the missing capability

Leading an AI initiative is a job, not a standing agenda item. Whoever holds it has to turn a broad ambition into a defined outcome, pull business and technical people in the same direction, weigh cost against risk and scope, and answer for whether the work is going anywhere.

That can sit internally when someone has the experience, the authority and protected time. All three matter. Give it to a capable manager without protected time and it becomes the task they pick up at six o’clock, which means it barely gets picked up.

Where those conditions are not in place, the capability can be brought in for a period.

There is a second reason to settle ownership early. When nobody internally holds the plan, procurement effectively starts on the supplier’s terms. A supplier should be answering a requirement you have written.

What good ownership looks like

Ownership changes how the work runs day to day. People know where decisions get made. Suppliers receive a requirement instead of a wish. Progress can be judged on something firmer than the number of meetings held.

A well-led initiative has most of the following:

• A named owner with the authority to decide and the time to do it.
• A defined business or service outcome. “Explore AI” is not one.
• A scope and sequence: what gets decided, what gets tested, what gets delivered, in what order.
• Specialist input where the knowledge is not in the building.
• Agreed roles for business teams, IT, risk, data owners and suppliers.
• Requirements written down before a platform or provider is chosen.
• A named owner for privacy, security, governance and the points where a person must review or approve an output before it affects a customer, an employee or an important decision.
• Measures covering whether the process has improved, whether people are using it and what it is worth.

This is not about taking decisions away from your people. It gives them somewhere to take them.

How to give AI progress the leadership it needs:

1. Name the outcome

“We should do something with AI” cannot be planned or measured. Name the operational, commercial or service outcome you want to improve. It might be the quote that takes three days to produce, the monthly reporting pack that absorbs a week, or the enquiries nobody reaches until Thursday. The owner needs something concrete to work on.

2. Put one person in charge

One name against the plan. A steering group can advise, but accountability shared across six people is accountability nobody feels.

3. Find out what is really happening

Before another pilot or another licence, ask who is using what. Tools bought on a card or accessed through a free account rarely appear on an IT asset list. Collect the pilots, the trials and the supplier conversations already underway into a single view, then decide what continues.

4. Fill the experience gaps

Buy in knowledge for the decisions your team has not had to make before. That normally means feasibility, data use, privacy and security, supplier terms, governance and delivery planning. You do not need that expertise permanently. You need it at the point the decisions are made.

5. Agree the need before you approach the market

Set out the process, the users, the data involved, the constraints and where someone must check an output before it goes any further. Then look at what is available. Options can be judged against your requirement rather than against each other’s demonstrations.

6. Make progress visible

Agree a short set of measures and a date to review them. The owner should be able to say what has been learned, what has been decided, what is stuck and whether the work should carry on. Stopping is a legitimate answer, and it is easier to reach when somebody owns the decision.

How AI Terrain leads the work when you have no internal capacity

AI Terrain is the AI consultancy trading name of Perform Partners. We provide the leadership, specialist knowledge and delivery structure that organisations without an internal AI team are missing. We can hold the work ourselves, or sit alongside an internal owner who has the authority but not the time or the AI-specific experience.

We start with the outcome and the way the work happens now, including the parts where your people’s knowledge and judgement are doing the heavy lifting. From there we test what is feasible, identify the controls and decisions required, and coordinate the people involved in delivery. We stay solution-agnostic, so what we recommend follows the business need and not a platform we happen to favour.

Depending on where you are, that might mean:

• leading the initiative and holding the plan, the decisions and the actions
• supporting an internal lead with structure, specialist advice and delivery capacity
• bringing separate pilots, ideas and supplier conversations into one prioritised plan
• assessing the opportunity before a tool or provider is chosen
• defining ownership, governance, human oversight and how value will be measured
• coordinating work across your teams, your IT function and your chosen suppliers

Sometimes AI is not the answer, or not the answer yet, and the honest recommendation is a process change or a decision to wait. We would rather say that early than help build a case for work that will not repay it.

How long we stay depends on what you already have. In some engagements we hold the work together from the start. In others we help an internal lead establish a workable approach, then step back as your own capability builds. Book a one-hour, senior-led transformation advisory session that helps you pressure-test a business, change, technology or transformation decision before committing budget or resource.

What to settle on before delivery starts

An owner can only lead when the important questions have answers. Before committing to delivery, you should be able to say:

• what specific outcome this work is meant to achieve
• who owns the plan and has the authority to move it forward
• which knowledge you hold internally and where you need outside help
• how work across teams will be coordinated
• what requirements suppliers will be assessed against
• who is accountable for data protection, security and human sign-off
• how progress, adoption and value will be measured

You do not need a large internal AI team to answer these. You need one person responsible for getting the answers and acting on them.

AI leadership and capacity: the questions leaders ask

Can you lead our AI initiative for us?

Yes. AI Terrain can lead the planning, decisions and coordination, or work alongside an internal owner who keeps the lead role. We agree the role, decision rights and responsibilities at the start, including which decisions and accountabilities stay with your organisation.

We have an IT team but no AI experience. Where do you fit?

Alongside them, not around them. Your IT team knows your systems, your controls and what will survive contact with your users. We add the AI-specific knowledge and the capacity to run the initiative, and they stay involved in the decisions that affect their estate.

How do we stop a supplier setting our direction

Write the requirement first. Define the outcome, process, users, data, controls and delivery constraints, then assess the options against them. We are solution-agnostic, and where a commercial relationship could affect a recommendation, we will tell you.

What if AI turns out not to be the answer?

Then we say so. Some problems are better solved by changing the process, fixing the data or removing a step, and you are entitled to hear that before you spend anything on tooling.

We are not sure we are ready. Is it too early to get help?

Buying a tool may well be premature. Getting clear on the problem is not. The Recce gives you an hour with a senior adviser to look at the ambition, the capacity you have and the most useful next move. That may be a fuller assessment, a piece of work your own team can do, or a decision to leave it for now.

Start with an honest conversation about where you are

Request a Recce: a free hour with a senior adviser, spent on your organisation and the problem you are trying to solve. You will get a straight view of where you stand, what deserves attention first and what a practical next step would involve.