If you only followed the latest announcements from OpenAI, Anthropic, Google, and other frontier AI companies, you would be forgiven for believing that enterprises are on the verge of a rapid transformation into autonomous, AI-native organizations. The narrative is compelling for the C-suite. AI agents could replace large portions of knowledge work; organizations might be able to dramatically reduce costs. Entire industries could be reinvented almost overnight.

I believe that story gets the direction right, but the timeline wrong. The distinction matters; if AI transforms enterprises over the next decade, organizations have time to learn, adapt, redesign operating models, as well as capture value. If those changes happen in the next 12 to 18 months, many businesses face profound disruption. Based on what I see across large enterprises, the evidence points much more toward the first scenario than the second.

The Headlines are Running Ahead of Reality

Many of the bold claims surrounding AI deserve healthy skepticism. That is not because the technology lacks capability; AI is unquestionably powerful, and frontier models continue to improve at an impressive pace. However, there is a significant difference between demonstrating a capability in a laboratory, and transforming an enterprise at scale.

Every few weeks, we hear that a new model can perform many of the tasks of software developers, lawyers, customer service agents, and more. There may be truth behind these announcements, but performing isolated tasks is a long way from reshaping an entire profession or operating model.

These announcements also serve another purpose. They reinforce investor confidence, and support the valuations of the companies building these models. Business leaders should be careful not to confuse marketing momentum with operational reality.

Enterprises are Making Progress, but It is Incremental

When I look across our client base, I do not see organizations standing still. I see hundreds of AI initiatives underway. The main takeaway is that most companies are focusing on improving existing processes, rather than reinventing them with AI tools. Executives frequently tell me they have invested heavily in AI tools for software development, customer service, or internal productivity. Token costs have increased substantially. Yet the productivity gains have generally been modest.

There are many reasons for this. Sometimes the tools themselves remain incomplete; in other cases, organizations have not changed the way they work. Simply giving employees AI assistants rarely delivers transformational results. In some instances, companies have actually needed more people, not fewer, to support the new workflows.

Over the past two years, we have consistently found that technology is only one part of the equation. Organizational redesign remains the much larger challenge.

The Biggest Barrier is Not the Technology

The limiting factor today is increasingly not the models themselves. Organizations still need to build operationally accountable AI systems that consistently deliver the right outcome at the right cost, and that remains difficult. At the same time, companies must rethink governance, workflows, incentives, training, and organizational structures. Those changes are inherently slower than deploying new software.

This explains why many organizations are moving cautiously. The risk of redesigning an operating model incorrectly often outweighs the near-term productivity gains. We are all experimenting, but experimentation should not be confused with near-term transformation.

Boards Need to Maintain Better Expectations

One consequence of the AI hype cycle is that expectations at the board level have become unrealistic. I regularly hear stories of CEOs visiting AI companies in Silicon Valley and returning with mandates to reduce costs by 40% through AI. That is an understandable aspiration; however, it is simply not what the evidence currently supports as a near-future reality.

Historically, organizations have achieved dramatic cost reductions through outsourcing, labor arbitrage, and operating model redesign. AI has not yet demonstrated the ability to consistently deliver those outcomes across enterprises. The better question is not whether 40% productivity improvements are possible; it is over what timeframe they can realistically be achieved. My expectation is that many of those gains will prove achievable, but over three to five years, rather than six to eighteen months.

Two AI Journeys are Happening Simultaneously

I increasingly see enterprises pursuing two distinct AI strategies. The first is an evolutionary path. Take software development as an example. Our research indicates organizations can achieve productivity improvements approaching 30% by applying AI to today's software development lifecycle. However, those gains require changing team structures, development processes, management practices, and more. Simply adding AI coding assistants will not deliver the full benefit. This evolution is likely to unfold over the next 18 months to two years, with further gains accumulating over a three to five year horizon.

The second path is reinvention. Rather than improving today's processes, organizations ask first-principles questions about what work should exist at all. Customer experience illustrates the distinction; an evolutionary approach focuses on reducing handling time, lowering cost to serve, and improving agent productivity. A reinvention approach asks an entirely different question: How can we resolve a customer's issue completely, on time, and in full?

Answering that question requires integrating customer service with sales, finance, fulfillment, marketing, and other adjacent functions. AI becomes the platform that enables a fundamentally different operating model rather than simply automating existing tasks. This second journey is significantly more ambitious and will likely take five years or more to mature.

Patience is Key for Maintaining Competitive Advantage

The organizations that benefit most from AI will not necessarily be those making the biggest announcements today. They will be the organizations that set realistic expectations, invest in organizational change alongside technology, and pursue both evolutionary improvements and longer-term reinvention simultaneously.

AI will undoubtedly reshape enterprise operations. I remain convinced of that. What I am less convinced about is the pace suggested by today's headlines.

For most organizations, the transformation has already begun. It is simply unfolding over years rather than months, and leaders who recognize that distinction will make better decisions, allocate capital more effectively, and ultimately, capture more sustainable value.