Expert interview with our CTO, Andrew Lawrie

The industry conversation rarely moves away from the subject of artificial intelligence presently. That’s why, we thought we’d corner our CTO, Andrew Lawrie to quiz him on his perspectives of where we are with AI, and how Workspend is approaching the subject of a Human-AI future.  

Digital complexity has become one of the biggest barriers to speed, visibility, and control. As businesses layer on more platforms, workflows, data sources, and automation tools, many are finding that more technology does not always create more clarity. In this interview, Andrew explains why (what he calls) “native AI” offers a more effective path forward — one that reduces friction, connects fragmented processes, and brings intelligence closer to where work actually happens.

Let’s start with the problem. What do you mean by digital chaos?

Digital chaos shows up when organizations have too many systems and data silos, too many disconnected workflows, and too little shared context. Teams spend time chasing information, reconciling versions of the truth, and working around the gaps between platforms. The problem is rarely a lack of technology. It is a lack of coherence.  

Why is this issue becoming more urgent now?

Most companies are managing hybrid teams, more specialised software, more external partners, and higher expectations around speed and responsiveness. That makes for a more complex operating environment.

Where have traditional digital transformation efforts fallen short?

A lot of transformation programmes focused on “ tasks” rather than redesigning how work will flow across the organization. That brought some efficiency gains, though it also created new silos. Many businesses ended up with modern-looking tools sitting on top of fragmented processes and data.

Why is native AI different – and what separates native AI from Co-Pilot AI?

Native AI is different because it is built into the operating fabric, not a bolt-on assistant in the margins of your screen. It’s “designed in” to data architectures,  workflows, and decision points. That means it can help interpret signals, reduce manual friction, surface next-best actions, and improve coordination in real time. Whenever AI is bolted on to incumbent systems, it often lacks (1) the organizational and role context, (2) the required content and (3) systems connectivity it needs to be useful.  This leads to hallucination and consequential impacts on user trust.  Native AI works because, as I previously mentioned, it’s “designed in.” 

How does native AI help reduce chaos in day-to-day operations in practical terms?

It helps in three ways. 

First, it reduces cognitive load by bringing together signals that would otherwise sit in different systems.

Second, it improves workflow discipline by guiding people toward the next action rather than leaving them to navigate complexity alone. 

Third, it strengthens visibility by giving leaders a clearer picture of what is happening across the operation.  

Where do you see the biggest gains from native AI?

Bluntly, “doing the jobs it’s good at!”  When it comes down to it, AI is a computer program (albeit a clever one) and, with it’s enormous data analytics and processing potential, it will always be good at processing data when appropriate conditions for it to do its work are in place.  We’ve had a bit of a false start with AI, because people have probably expected too much as the result of vendor hype.  It REALLY CAN do remarkable things when it comes to optimizing data processing and reporting work, but – like all of us humans – it needs to be given the right raw materials and ingredients to perform at its best.

What role do humans play in an Human-AI-native model?

A central one. AI should strengthen human judgement, not replace it. So far as we’re concerned, certainly at this stage in the maturing of AI tech, humans must always be “in the loop” as a necessary safeguard.  People still provide context, prioritization, ethical oversight, and relationship management. These are core to how good organizations work. Native AI works best when it takes care of repetitive effort and information friction, leaving people to focus on the decisions and interactions that carry weight.

What should business leaders do first if they want to move in this direction?

Start with the friction, not the technology. Look for the points where work slows down, handoffs break, visibility disappears, or teams duplicate effort. Then ask what data, workflow, and decision logic sit behind those moments. Native AI works best when it is applied to high-friction, high-value processes. The goal is not to add more tools. The goal is to create a more coherent operating model.

What is the biggest misconception about AI in the enterprise today?

One misconception is that AI value comes mainly from automation. In reality, a great deal of value comes from coordination. Organizations often struggle less with doing individual tasks and more with connecting actions across teams, systems, and time. That is where native AI becomes especially powerful. It helps turn fragmented activity into a more unified operational rhythm.

If you had to leave readers with one message, what would it be?

Digital chaos is more an operating model than a technology challenge … there is plenty of clever tech out there! Native AI matters because it gives organizations a way to move forward with AI in a safe, worthwhile, affordable, and results-driven way.  And yet, I would argue the real opportunity is not simply doing more with AI. It is creating a business that works with greater clarity, speed, and coherence.