The future of AI expertise: Building and managing AI-capable teams

As organisations adopt artificial intelligence (AI) more widely, a critical challenge emerges: how do you build and manage teams capable of delivering on AI’s promise of increased productivity, enhanced customer experiences, accelerated innovation, and sustainable competitive advantage?

In my formative years at AWS, I created the early thinking around designing a Cloud Centre of Excellence (CoE) and incorporating it into IT teams who were adopting the cloud. Now, as organisations embrace AI, I’m seeing familiar patterns and also crucial differences that demand a new approach.

Learning from the past

As cloud adoption picked up pace in our customers, I helped them establish Cloud CoE’s, typically within their IT departments. These CoE’s primarily focused on technical excellence and standardisation. This made sense at the time - cloud was primarily viewed as an infrastructure story, albeit one with significant innovation and business value potential.

But AI is different. While both cloud and AI are transformative technologies that face skills gaps and initially require centralised expertise, AI can’t be delegated to IT. AI touches every aspect of business operations and it is this comprehensive impact that demands top-down adoption and board-level engagement from day one.

The democratisation of AI has followed a fundamentally different path too. Where cloud tools took time to reach beyond technical users, AI has experienced a mass consumerisation moment. Tools like ChatGPT have launched AI directly into boardrooms everywhere, bypassing the traditional technology adoption cycle we saw with cloud.

Your AI CoE must sit with the Board

Let me be absolutely clear: your AI Centre of Excellence must sit alongside the Board, not within IT or your cloud CoE. This isn’t just a recommendation, it’s a necessity driven by the unprecedented risks and opportunities AI presents.

Organisations are moving from hundreds of decisions per day made by humans to millions of decisions per second made by AI. This exponential increase in decision-making velocity creates unprecedented governance challenges.

In my speaking engagements, I often show the audience this slide depicting directors in prison - it’s provocative, but it makes a crucial point. Without proper Board-level governance of AI initiatives, directors could face personal liability for inadequate oversight of AI systems making countless decisions that expose the organisation to risk.

The AI CoE needs Board-level placement because:

1. It directly shapes business strategy, unlike cloud which was primarily an enabler
2. It requires oversight of ethical, legal, and risk implications that far exceed technical concerns
3. It needs to drive cultural change across the entire organisation
4. It has to coordinate among diverse stakeholders from all parts of the business

Who are the AI stakeholders?

One of the most common misconceptions I encounter is that AI transformation involves the same stakeholders as cloud transformation. This couldn’t be further from the truth. Where cloud primarily engaged IT leaders, infrastructure teams, and development groups, AI demands engagement from:

• Board Members and Executive Leadership
• Business Unit Leaders
• Domain Experts
• HR (for workforce transformation)
• Ethics and Compliance Teams
• Legal Teams
• Line Managers
• End Users across the organisation

This broader stakeholder landscape reflects AI’s position as a business tool rather than just a technical capability. It’s why I frequently see organisations struggle when they try to apply cloud-era stakeholder engagement models to AI transformation.

Build your team for the AI era

Cloud success often came down to technical configuration and implementation skills. Get the setup right, and you were good to go. AI demands a fundamentally different skill set.

It requires people who are willing to ask more questions, and not just accept outputs. This applies not only to generative AI but to machine learning predictions as well. Teams need domain experts who can challenge results, demand explanations, and understand how AI reaches its conclusions.

This questioning mindset needs to be built into your organisational culture. It’s not enough to have technical expertise - you need people who can:

• Challenge AI outputs based on domain knowledge
• Understand and validate AI’s decision-making processes
• Identify potential biases or errors in AI systems
• Ensure AI decisions align with business objectives and ethical requirements
• Provide training inputs for AI models

Common pitfalls to avoid

In my work with Boards on AI transformation and IT teams on cloud adoption, I’ve observed several critical pitfalls:

• The “AI is just cloud+” Trap: Many organisations assume AI is simply an extension of cloud computing and therefore think they have the necessary structures in place. This misconception leads to inadequate governance and improper placement of AI initiatives within the organisation.

• The Technical Focus Trap: Over-emphasising technical skills while undervaluing domain expertise and critical thinking capabilities. Success in AI requires both technical competence and business understanding.

• The Governance Gap: Failing to establish proper Board-level governance from the start. This isn’t just about oversight, it’s about protecting your organisation and its directors from significant risks.

A framework for success

Throughout this article, we’ve explored why AI teams need different treatment from traditional technology teams and where they should sit in your organisation. But the practical question remains: how do you actually build and manage these teams?

Working with organisations across different stages of AI adoption, I’ve seen first-hand how the challenge of building AI-capable teams touches every aspect of organisational design and development. It begins with fundamental questions about team composition - not just what roles you need, but how to balance technical AI expertise with domain knowledge, and how to find the right mix of internal and external talent.

The management challenge is equally complex. Leading AI teams requires new approaches to performance measurement and team structures. How do you effectively orchestrate groups that combine AI specialists with domain experts? What does good performance look like when you’re blending human and machine capabilities?

Skills development becomes particularly crucial in this context. Traditional training programs often fall short when preparing teams for AI-enabled work. You need frameworks that develop both technical competencies and critical thinking abilities, while creating clear career paths that keep pace with rapidly evolving technology.

The talent question extends beyond just hiring. Yes, you need a strategy for external recruitment, but the often-overlooked opportunity lies in identifying and developing internal talent. Your existing domain experts, equipped with the right AI capabilities, can become your most valuable team members.

Cultural transformation underpins all of this. Moving from traditional teams to AI-capable ones isn’t just about skills - it’s about fundamentally changing how people think about their work and their relationship with technology. This shift requires careful planning and expert guidance to manage effectively.

Finally, there’s the critical question of governance and risk management. How do you establish effective oversight without stifling innovation? What controls need to be in place to manage AI decision accountability? These questions need answers before you scale your AI capabilities.

About the Author

Mario Thomas is a Chartered Director and Fellow of the Institute of Directors (IoD) with nearly three decades bridging software engineering, entrepreneurial leadership, and enterprise transformation. As Head of Applied AI & Emerging Technology Strategy at Amazon Web Services (AWS), he defines how AWS equips its global field organisation and clients to accelerate AI adoption and prepare for continuous technological disruption.

An alumnus of the London School of Economics and guest lecturer on the LSE Data Science & AI for Executives programme, Mario partners with Boards and executive teams to build the knowledge, skills, and behaviours needed to scale advanced technologies responsibly. His independently authored frameworks — including the AI Stages of Adoption (AISA), Five Pillars of AI Capability, and Well-Advised — are adopted internationally in enterprise engagements and cited by professional bodies advancing responsible AI adoption, including the IoD.

Mario's work has enabled organisations to move AI from experimentation to enterprise-scale impact, generating measurable business value through systematic governance and strategic adoption of AI, data, and cloud technologies.