There is a procurement arms race underway in organisations of every size and sector, and AI is at its centre. Boards are setting AI investment targets. Procurement teams are fielding an expanding list of AI vendors. Chief Technology Officers are under pressure to demonstrate AI ambition. And so organisations are buying: buying AI tools, AI platforms, AI consultants, AI training, AI governance software, AI strategy frameworks.
Much of this spending is producing very little value. Not because the products are bad — many of them are genuinely impressive — but because the organisations buying them have not done the thinking that would allow them to deploy them well.
"AI is not a product you can procure your way into. It is a capability you have to build — and building it starts with being honest about what problem you are actually trying to solve."
When AI adoption is driven primarily by procurement decisions rather than strategic clarity, a predictable set of problems follows. Tools are purchased before use cases are defined. Pilots are launched before the organisation has established how success will be measured. Vendors promise transformation; organisations deliver incremental improvement at best, expensive complexity at worst.
More subtly, procurement-led AI creates a dependency mindset. Instead of building internal capability to think clearly about AI — to identify problems worth solving, to evaluate approaches critically, to govern deployments rigorously — organisations outsource this thinking to vendors whose interests are not always perfectly aligned with their own.
To think well about AI, organisations need to be able to do three things that are deceptively simple to describe and surprisingly hard to do.
Identify problems worth solving. Not "where could we use AI?" but "what are the most consequential inefficiencies, bottlenecks, risks, or opportunities in our business, and which of them might AI help address?" This requires deep business knowledge and honest diagnosis — neither of which can be purchased from a vendor.
Evaluate approaches with rigour. AI is not a single technology. It is a broad family of techniques, each with different strengths, weaknesses, costs, and risks. Choosing well requires understanding the trade-offs — not at a technical level, but at the level of: will this actually work in our context, and what happens when it doesn't?
Build the organisational capacity to learn. AI systems require ongoing attention. They drift. They fail in unexpected ways. They create new ethical dilemmas that weren't anticipated at design time. The organisations that use AI well are the ones that have invested in the people and processes needed to monitor, adjust, and improve their systems over time — not just to deploy them and move on.
The organisations that are getting the most value from AI are not necessarily the ones that moved fastest. They are the ones that thought first. They spent time on strategy before they spent money on technology. They built internal capability before they scaled external deployment. They were honest about what they didn't know, and they invested in learning before they invested in doing.
This is not an argument for moving slowly. It is an argument for moving deliberately. The organisations that move fast without thinking first are not winning the AI race — they are accumulating technical debt, governance debt, and strategic confusion at speed.
DeepSlate helps organisations think before they buy — and act with confidence when they do. See our AI Strategy Sprint or get in touch.