In the first two articles of this series, I’ve explored why boards need an AI Centre of Excellence (AI CoE) and detailed the eighteen functions that determine AI success. But before you can build effective governance, you need to understand where you actually are today - not where you think you are, or where you’d like to be. To help boards navigate this challenge, I’m sharing the AI CoE Simulator - a practical assessment tool taken from my AI governance toolkit that operationalises the AISA framework for real-world use.

Every AI Centre of Excellence (AI CoE) needs a clear operational mandate. Through my experience designing and building Cloud Centres of Excellence for AWS customers, extensive research, and practical implementation, I’ve identified the essential functions that provide comprehensive AI governance without creating bureaucratic overload. These functions, organised around the Five Pillars mechanism, ensure your AI CoE can effectively govern multi-speed adoption while building the capabilities needed for sustainable AI transformation. Understanding these functions, and how they interconnect is crucial for boards establishing effective AI governance.

Boards face an unprecedented governance challenge as AI systems operate at speeds traditional oversight cannot match. With the vast majority of AI pilots failing to reach production and shadow AI creating unmanaged risks across organisations, establishing an AI Centre of Excellence has become essential board infrastructure. This article explores how an AI CoE provides the governance framework needed to coordinate multi-speed AI adoption, build capabilities systematically across the Five Pillars, and transform ungoverned AI risk into strategic competitive advantage.

I’m regularly asked how to use the AI Stages of Adoption, Five Pillars, and Well-Advised together practically. In this article I explain how these three mechanisms integrate to address the unique challenge of AI’s multi-speed adoption across different business functions. I provide a straightforward approach for boards to coordinate AI transformation whilst managing the governance complexities that emerge when different parts of the organisation advance at different speeds.

Even the most meticulously crafted AI business case can fail at the final hurdle - securing Board buy-in. With research showing 88% of AI pilots never reach production, effective presentation isn’t just about gaining initial approval but establishing the path to full implementation. This final article in my series explores how to present AI investment proposals to Boards, addressing their six key areas of concern while building the stakeholder confidence necessary for successful transformation. By understanding Board dynamics, anticipating objections, and structuring presentations that balance strategic vision with implementation rigour, you can navigate the critical journey from business case to production-scale AI.

Organisations are demanding disciplined, comprehensive business cases for AI initiatives that balance traditional financial rigour with frameworks capturing AI’s unique value creation patterns. In this fourth article in my series on AI business cases, I provide a step-by-step guide to building AI business cases that secure approval and set the foundation for successful implementation.

Finding high-value AI opportunities requires looking beyond the obvious. While most organisations gravitate toward trendy applications like chatbots, the most impactful AI initiatives often lie in less visible but more strategically significant processes. By applying a structured evaluation approach that examines process characteristics, strategic alignment, and implementation feasibility, boards can identify AI investments that deliver transformative value across multiple business dimensions. This systematic method ensures scarce resources target opportunities with the greatest potential impact rather than those with merely the highest visibility or short-term appeal.

Building on my previous thoughts about why traditional business cases fail for AI investments, this article explores what I consider to be the essential building blocks for a more effective evaluation approach. This foundation provides boards with the tools to assess AI’s unique value creation patterns while maintaining financial discipline - helping leaders confidently navigate investment decisions that conventional models simply cannot adequately evaluate.

In today’s AI-driven landscape, traditional business case methods fall short when evaluating AI investments. Drawing from my experience developing AWS’s cloud business case tools, I explore why conventional ROI models struggle with AI’s parallel, multi-speed adoption patterns. Unlike cloud’s sequential journey, AI initiatives exist simultaneously across different maturity stages, creating valuation challenges that standard metrics can’t capture. Boards need new evaluation approaches that account for AI’s diverse cost structures, varying timelines for returns, and how investments in one area often enable value in entirely different parts of the business.

As I discussed in my article on building and managing AI-capable teams, organisations face a critical challenge in acquiring the right talent for AI transformation. This reminds me of the early days of cloud adoption, when I advised enterprises on their migration strategies. Back then, I witnessed the same scramble for scarce talent, which led me to advocate strongly for upskilling existing teams rather than relying solely on external hiring.