“AI does not create value”

On the surface, this statement appears to be indefensible. Around the world, companies are injecting billions of dollars into AI precisely because they expect to see value creation—increased efficiency, smarter decisions, innovative products.

Yet, the claim holds an important truth.

AI does not create value merely by existing. The same is true for a new software platform, a faster database, or modern cloud infrastructure. Technology provides capabilities. Value only materializes when those capabilities result in the transformation of a business outcome.

Often this distinction gets overlooked.

Many AI projects are benchmarked on technical performance. Teams evaluate model accuracy, the quality of an AI’s response, its latency, the speed at which it can be deployed, or how many users adopt it. These are excellent metrics of an engineering project’s success; they prove a solution has been successfully built and deployed in alignment with the design.

However, these measures do not necessarily indicate business impact. A chatbot that accurately answers a customer’s question does not inherently lower support costs. A model that demonstrates outstanding accuracy on benchmark tests does not automatically optimize inventory.

An AI coding assistant that a developer finds helpful may not inherently reduce cycle time or boost revenue.

The missing piece is the business outcome. Strongest AI initiatives often start by establishing the business outcome.

Which metric should be influenced?

How will we define success?

What financial or operational outcome would justify the investment in this technology?

Only once these are clearly defined should the organization begin considering whether AI is the right tool for the job. By approaching projects with this perspective, the nature of the discussion completely shifts. Instead of the conversation centering on, “Can we build this with AI?,” the organization starts with, “What measurable improvement are we seeking to achieve?”

The difference between these questions may seem small, but it fundamentally reshapes priorities. This also helps explain why certain AI projects fail to deliver the expected results despite excellent engineering. Engineers will always naturally try to optimize the product they build: the model, the pipeline, the prompts, the interface.

These are concrete, testable quantities.

Value to the business, on the other hand, resides beyond the system’s boundary—it emerges when the organization reconfigures its processes or strategy to improve a metric that matters. This is why organizations increasingly want technical talent to be conversant in business metrics. It is no longer enough to simply build the right technology; decision-makers want to understand the impact it has. Did the AI drive operating margin up, cut handling time down, or lift conversion rates?

These outcomes carry more weight than the choice of architecture.

Looking at the world from this lens, the claim that “AI does not create value” is both true and false. It is false because the technology clearly has immense economic value. It is true because that value is not inherent in the technology itself; rather, it’s the result of applying the technology to improve outcomes and shape a business’s future.