The first steps in managing AI as an asset – Transcript
In this video series, I’ve been making the case that we should manage AI as an asset—a strategy that goes beyond controlling costs and risks to genuinely unlocking innovation, clearing barriers to effective usage, and enabling true differentiation in the marketplace. We’ve discussed the nature of AI sprawl in the enterprise, its associated risks, the barriers to adoption, and the looming regulatory challenges. Today, I’d like to focus on the practical first steps we can take to govern AI properly within our organisations.
Before we dive in, let’s briefly recap the key risks: we’ve covered security and privacy, unchecked spend, and regulatory pressure. There’s also the matter of “grey AI”—where employees, keen to innovate, bypass official channels. This can lead to unapproved data flows, inappropriate usage of SaaS-based AI tools, and serious compliance breaches. For example, I’ve heard of instances where well-intentioned staff fed sensitive customer data into a public AI platform, inadvertently triggering potential lawsuits. It’s a stark reminder that even the brightest innovators need guidance and guardrails.
So, where do we start when it comes to managing AI like an asset?
Discovery:
Begin by establishing a process for identifying which AI tools, systems, and use cases are present in your organisation. From high-risk applications to those on the lower end of the scale, you need a clear picture of what’s out there before you can shape governance effectively.
Classification & Use Cases:
Once you’ve identified these systems, classify them by their business purpose, level of risk, and potential impact. Understanding the “what” and “why” of each AI solution will inform how you oversee and support it.
Ownership & Accountability:
Like any valuable asset, AI needs someone at the helm. Assign clear system owners—especially for high-risk scenarios—and define their responsibilities. Establish well-communicated guardrails so innovators know what’s permissible and how to experiment without jeopardising the organisation.
Financial Control & ROI Considerations:
AI involves substantial infrastructure demands—think high-performance compute, significant storage, and increased cloud consumption. This can rapidly inflate costs. By managing AI as an asset, you can better track spending, prevent runaway bills, and ensure initiatives deliver tangible value. Not every promising use case justifies the cost; ROI must remain front and centre as you scale and refine these solutions.
If all of this seems daunting, remember that we’re still in the “Wild West” phase of enterprise AI. We’re at the ground floor of what will surely become a more structured, tool-rich environment. Just as we’re seeing bad actors and questionable usage emerge, we’re also witnessing dynamic innovation—new capabilities aimed at discovering, monitoring, and governing AI systems. There’s no doubt the technology landscape will mature, offering more effective tools and best practices that make this journey easier.
In the next and final video, I’ll bring all the threads of this series together, distilling them into actionable steps and takeaways. We’ll look at what you can do right now to manage AI as an asset and lay the foundation for a successful, sustainable AI strategy within your organisation.
View the video on Youtube here: https://youtu.be/SiO5GYcEync