Skip to content

Get the monthly AI briefing: what matters for your business, in plain English.Subscribe

Menu

AnalysisAI GovernanceBuilding Capability

The evolving role of the AI leader

AI started as an IT project. Now it is a question of strategy, people and accountability. How AI leadership is changing and what the role must cover.

Phil Evans

Co-Founder & CEO

For most businesses, AI began as an IT project. It is now a question of strategy, people, risk and accountability, and the answer to “who owns AI?” has changed with it.

This article looks at how AI leadership is changing, what UK law and regulators expect, the main ownership models and what the role now needs to cover.

From IT project to leadership question

The evidence shows AI moving out of the technology function and into the wider business.

  • The money has moved. BCG’s Applied AI Index, published in September 2026, found that AI spending had risen from about 1.7% of revenue in late 2025 to 3.3%, and that more than 80% of it now sits outside enterprise IT.
  • New roles are appearing fast. IBM’s 2026 study of 2,000 chief executives found that 76% of their organisations now have a chief AI officer, up from 26% a year earlier. These are large organisations, but the direction is clear.
  • Every leader is becoming an AI leader. In the same study, 85% of respondents said every functional leader should become a technology expert in their own domain, and 79% of executives said they are decentralising decision-making.
  • AI is starting to make decisions. Chief executives in the IBM study expect AI to make 48% of codifiable operational decisions without human intervention by 2030, up from 25% today.

Most smaller businesses will never appoint a chief AI officer. They face the same question all the same: who decides how AI is used here, and who answers for the result?

Accountability does not move with the technology

Whatever role a business creates, UK law and regulators keep accountability where it already sits.

  • Directors. Directors’ duties under the Companies Act 2006, including the duty in section 174 to exercise reasonable care, skill and diligence, make no exception for decisions about AI.
  • Financial services. In October 2026, the FCA published its review of good practice in AI. It describes firms that deploy AI safely as having senior managers who are accountable for AI systems and take reasonable steps to mitigate the risks. It highlights firms that mapped AI responsibilities across their senior managers and assigned accountability from the design stage, and flags poor audit trails as an area for improvement. The FCA and PRA have told the Treasury Select Committee, whose report on AI in financial services was published in January 2026, that they did not think a dedicated senior manager role for AI was needed: accountability sits with the senior managers whose areas AI affects. The review also flagged senior managers who had handed responsibility for AI entirely to technical teams.
  • Professional services. In Ayinde v Haringey (June 2025), the High Court said that those with leadership responsibilities, such as managing partners and heads of chambers, must take practical and effective measures to ensure that everyone providing legal services understands their obligations when using AI.

The practical conclusion is simple. A business can appoint someone to lead on AI, but it cannot delegate the accountability that comes with running the business.

Where boards are today

  • AI is reaching the agenda, slowly. Deloitte’s 2025 survey of 695 board members and executives in 56 countries found that 31% said AI was not on their board’s agenda, down from 45% in its previous survey. 66% reported that their boards had limited or no AI knowledge or experience, down from 79%.
  • UK directors recognise the gap. In a 2025 Institute of Directors survey of more than 600 UK directors, half named limited experience among managers and board members as a top concern about AI.
  • Ownership at the top is still the exception. McKinsey’s survey published in March 2025 found that only 28% of respondents said their CEO was responsible for overseeing AI governance, and 17% said the board was. It also found CEO oversight to be one of the factors most correlated with bottom-line impact from generative AI.
  • Regulated firms are further ahead. The Bank of England and FCA’s 2024 survey found that 84% of financial services firms using AI had an accountable person for their AI framework.

Four ways to organise AI ownership

There is no single right model, and the FCA is explicit that there is no one-size-fits-all approach. Most approaches are a version of one of these four.

  • Senior executive owner. How it works: AI becomes part of an existing executive’s role, such as the managing director or chief operating officer, with specialist support. Often suits: Smaller and mid-sized businesses. Watch for: Time and expertise. The owner needs support and a clear mandate, not just a title.
  • Dedicated AI leader. How it works: A chief AI officer or head of AI coordinates strategy, value and risk across the business. Often suits: Larger or AI-intensive organisations. Watch for: The role becoming a silo, and other leaders treating AI as someone else’s job.
  • Technology-led. How it works: AI sits with the chief technology, information or data officer. Often suits: Businesses where AI is mainly internal tooling. Watch for: Value, people and conduct questions being underweighted, when most AI spending now sits outside IT.
  • Federated. How it works: A small central team sets standards and controls, and each function owns its own uses. Often suits: Organisations with many functions using AI. Watch for: Inconsistency, unless the centre is strong enough to set and enforce common rules.

Whichever model you choose, the FCA’s review points to a useful refinement: name an owner for each use of AI, with one senior person accountable for oversight across all of them.

What the AI leader now has to cover

The job has grown well beyond choosing tools. It now spans six areas.

  • Value. Deciding where AI should be used and measuring whether it is working.
  • Risk and compliance. Knowing which rules apply, from UK GDPR to sector regulation and, where relevant, the EU AI Act, and making sure they are met.
  • People. IBM’s 2026 study found that only 25% of the workforce uses AI regularly, and 83% of chief executives said AI success depends more on adoption by people than on the technology. Skills, training and change are now central to the role, and 59% of chief executives expect the influence of the chief HR officer to grow.
  • Suppliers and data. Understanding which providers the business depends on and what happens to its data.
  • Decision rights. As AI takes on more decisions, someone has to decide which decisions it may make, with what oversight, and how they can be reversed.
  • Evidence. Recording what was decided, why and who challenged it. Without a record, it is hard to show that reasonable steps were taken.

This is why the role increasingly calls for a translator rather than a technologist: someone who can connect commercial goals, risk and people, who has credibility with the board and who is willing to say no.

Practical steps

  1. Name one accountable executive. Even in a small business, one senior person should own AI. A shared responsibility tends to become nobody’s.
  2. Map who does what. For each significant use of AI, record who decides to use it, who builds or buys it, who checks the output and who signs off.
  3. Assign ownership at the start. Give each new AI use an owner before it is built or bought, not after it is in use.
  4. Put AI on the board agenda. Agree what the board should see and how often: use, value, risk and incidents.
  5. Build board and leadership literacy. Short, practical briefings help directors challenge proposals with confidence. For businesses within the EU AI Act’s reach, the AI literacy duty applies here too.
  6. Decide decision rights explicitly. Write down which decisions AI may make on its own, which need human approval and which it must never make.
  7. Keep the record. Document decisions, the reasoning and the challenge. It is the simplest way to show that leadership is in control.

AI leadership is less about who understands the technology best than about who is answerable for what it does.

Sources checked October 2026. Several surveys cited are of large organisations and may not reflect smaller businesses. The FCA’s good practice examples do not create new requirements. Regulatory expectations may change and should be re-checked before reliance.

This article provides general information and does not constitute legal or regulatory advice.

References

Research on AI leadership

Boards and directors

Regulators and courts

ShareLinkedInEmail

Keep learning

AnalysisAI Governance

The business case for AI governance

AI governance is usually filed as a compliance cost. The evidence says it is an investment, and that the cheapest time to make it is now.

Phil Evans

Co-Founder & CEO

Building Capability

Put it into practice

The AI Jargon Buster

Starter PackAI Adoption

The AI Jargon Buster

Twenty words you will hear in every AI conversation, in plain English. What each one actually means, and a way to picture it

Isaac Ward

AI Practitioner · 14 Oct 2026

Free downloadGetting Started

Stay informed

A short monthly briefing on the AI developments that actually matter for your business. No sales pitches.

Monthly. Unsubscribe in one click. We never share your details.