At recent KonnectHouse Agentic AI events in London and New York, a consistent theme emerged: everyone is using and building AI, but few are clear on what actually differentiates it. The familiar “build vs. buy” debate has also resurfaced—now sometimes reframed as “build, buy, or adopt.” But even that evolution misses the point. The real question is no longer where AI comes from. It’s what solution or set of solutions will procurement use to govern it.

Access to AI Is Becoming Standardized

Access to AI capabilities is becoming standardized across the market. Whether embedded in Coupa, SAP, Ivalua, or deployed through established best of breed, or nascent AI-first Procuretech, most solutions are leveraging the same underlying models.

The artificial intelligence layer itself is no longer a durable source of differentiation—it is rapidly becoming table stakes. What is not standardized is how AI is applied “within the enterprise”—how outputs are made relevant, embedded into workflows, and translated into consistent, coordinated actions.

At the KonnectHouse event in NYC, solution differentiation using AI took on many forms. Intake and orchestration platforms positioned themselves as the new enterprise coordination layers. Domain-specific providers emphasized their use of AI to transform specific functions with functional depth in areas like sourcing and contract management. Large suites leaned in on scale, system-of-record advantage while boasting the acquisition of smaller AI players into their existing suites.

Different narratives—but all had the same dependency on a shared or similar AI infrastructure that is rapidly evolving.

The Build vs. Buy Debate Is Being Replaced by a New Reality

Given the pace of innovation, procurement teams are no longer simply choosing between suite-based solutions and best-of-breed vendors. They are increasingly building their own micro applications within AI platforms—using natural language to generate, refine, and debug code—reducing their reliance on traditional coding expertise and IT-led development.

Standalone LLMs—such as those offered by Anthropic, OpenAI, Meta, and Google—are now widely accessible and increasingly capable. They can already without much procurement context, analyze spend, generate sourcing events, and recommend suppliers with minimal setup.

This has reduced the barrier by potentially empowering procurement teams to create their own functional AI-powered intelligence and workflows in days—not months. In some cases, procurement teams are already building intake flows, supplier evaluation agents, and sourcing support tools internally. This changes the dynamic entirely and this is where the real shift is happening.

The Pressure on Traditional “Procurement” Differentiation

Moreover, this shift is already creating pressure for both vendors and enterprises. For procurement domain-focused providers, the traditional argument—that specialized tools win because they understand procurement—is weakening. Standalone LLMs are rapidly closing the gap on functional knowledge, making capabilities like contract analysis, sourcing optimization, and spend classification widely accessible.

For enterprises, the economics are shifting as well. When internal teams can replicate meaningful portions of functionality quickly, the value of traditional SaaS models comes under pressure. Functionality is no longer enough. The question becomes: how does a solution fit into—and help govern—a broader system of decisions and actions?

So the risk is no longer a single bad decision. It is the proliferation of many independently built agents managing workflows—each locally optimized, but globally misaligned. This is the emergence of “shadow orchestration.” And unlike shadow spend, it operates at machine speed.

The Real Divide Is Just Starting to Emerge

The conversations at KonnectHouse made one thing clear: the market is converging faster than it appears. The distinction between AI-first startups and established suites is narrowing. The difference between building, buying, or adopting is becoming less meaningful as capabilities standardize.

The winning solutions will not simply embed AI. They will define a role within a broader system—owning critical workflows, acting as orchestration layers, or embedding directly into enterprise decision-making in a way that ensures consistency and control. In a more autonomous environment, alignment—not capability alone —will define differentiation.

Where Liberis Is Focused

In a world where discussions around AI are no longer scarce, solution differentiation will not come from AI models and agents alone. This is the shift we are actively helping solution providers navigate at Liberis Consulting. The question we hear most often is simple: “If everyone now has AI, how do we stand out?”

Learn more about how we can help answer this at Liberis Consulting.