Also called: AI CoE · Last checked: 7 November 2026
In short: An AI centre of excellence (AI CoE) is a central team of AI specialists. It sets standards, shares tools and know-how, and helps the rest of the business build and use AI safely.
Table of Contents
How it works
- Gather the experts. The business brings data scientists, engineers, and risk and ethics specialists into one team.
- Set the standards. The team writes rules, picks approved tools and builds shared data sets and models.
- Support the business. Departments bring ideas, and the centre helps them test, build and run AI tools.
Where you’ll meet it
Banks, insurers, retailers and public bodies often announce an AI centre of excellence when they hire a head of AI.
What it means for your business
- Start small. Two or three people with a clear brief can set rules and support early projects.
- Run it as a service. Measure the centre by how many business teams it helps get AI tools into daily use.
- Plan its end state. As skills spread, the centre often shrinks to standards and oversight.
Watch out for
- Distance from the work. A centre that rarely talks to front-line teams builds tools nobody uses.
- Bottleneck. If every project must queue for the centre, teams stall or work around it.
Example
Lloyds Banking Group. In August 2024 Lloyds appointed Rohit Dhawan, a former Amazon Web Services executive, as its first group director of AI and advanced analytics. He was hired to lead the bank’s AI Centre of Excellence, which brings together data scientists, behavioural scientists, machine learning engineers and AI ethics experts. The centre already included more than 30 PhDs, Lloyds said.
AI centre of excellence vs AI operating model
| AI centre of excellence | AI operating model | |
|---|---|---|
| What it is | A central team of AI experts | The whole set-up for running AI across the business |
| Covers | Standards, expert support and shared tools | Structure, decision rights, funding, rules and skills |
| How they relate | One part of an operating model | May include a centre of excellence |
Frequently asked questions
How big should an AI centre of excellence be?
A mid-sized firm can start with two to five people. Large firms may have dozens, but the number of teams the centre helps matters more than its size.
Who should lead an AI centre of excellence?
Someone who combines technical knowledge with business sense. They need enough authority to set standards and enough contact with departments to know their real problems.
Does an AI centre of excellence replace the IT team?
No. IT keeps running systems, security and support. The centre focuses on AI methods, standards and helping departments use AI well.