Next AI race shifts from building to deployment

The race to develop artificial-intelligence technology will continue. But as AI becomes more available, the ability to apply it effectively and reorganize businesses around it will become increasingly important.

In my view, this will define the next phase of the AI era. The first phase has largely been about creating increasingly powerful intelligence. The next phase will be about how companies put that intelligence to work.

From building intelligence to deploying intelligence

The real difference will emerge when companies redesign the way they do business around AI, enabling business models, customer relationships, supply chains and organizations to operate differently. The distinction will increasingly lie in how deeply companies can translate AI capabilities into actual business capabilities.

But there is an important difference between using AI within an existing organization and changing the organization because AI now exists. Most companies are still doing the former. However, the larger transformation lies in the latter.

Technology is moving faster than organizations

One way to understand this challenge is to look at three different dimensions of AI transformation: technology, talent and organization.

Technology moves the fastest. New models and applications appear almost every week, and capabilities that seemed experimental only months ago are quickly becoming practical.

Talent follows. People gradually move from using AI as a search or writing tool to delegating more sophisticated tasks, evaluating its output, and incorporating it more deeply into their work.

Organizations usually move the slowest. A company may have been built over 20, 50 or even 100 years. Its departments, reporting lines, workflows, incentives and management layers reflect assumptions accumulated over a long period of time. Those assumptions do not disappear simply because a new AI model has been released.

Many of these structures were designed for a world in which information was difficult to obtain, expertise was concentrated among specialists, and coordination depended heavily on managers. AI begins to loosen some of these constraints.

The result is a growing gap between what AI makes possible and what the organization is able to absorb. A company may equip its employees with powerful AI tools, but if information remains fragmented, decisions still move through multiple layers and workflows remain unchanged, much of AI’s potential will remain trapped inside the old organization.

In my view, this will become one of the central challenges of AI transformation.

Four stages of AI transformation

At the AI-aware stage, companies recognize the importance of AI and begin experimenting. AI helps people perform existing tasks better.

As companies become AI-active, adoption spreads across functions, supported by common platforms, governance mechanisms and clearer organizational responsibilities, building the capabilities to deploy AI at scale.

The more important transition comes when a company becomes AI-driven. Companies begin to reconsider the work itself, moving from improving existing processes to reshaping them.

Ultimately, some companies will move toward becoming AI-native. Human employees and AI agents work together, knowledge can move more easily across the organization, and some decisions can be made closer to where the work actually happens.

What happens when the workflow changes

If AI produces analyses faster while the same layers of review and decision-making remain in place, the underlying process has changed only marginally.

Companies will still require direction, accountability and coordination, and industries will evolve at different speeds.

A smaller human team supported by specialized AI agents may increasingly be able to handle work that previously required a much larger group. The significance goes beyond efficiency or labor substitution: It could gradually change how companies think about team size, roles and organizational design.

The rise of human-AI teams

Instead of using one general-purpose assistant, employees may work with several agents that perform different roles. Some employees may lead a combination of human colleagues and AI agents.

Experience will certainly remain important, but access to knowledge is becoming much less scarce. As this becomes more common, the value of people will shift increasingly toward what AI cannot easily provide on its own: defining the real problem, exercising judgment in ambiguous situations, creating new possibilities, and challenging outputs that appear convincing but may be wrong.

Leadership will change for similar reasons. Some decisions that once required escalation may be handled closer to the customer or the operation. Managers’ roles will move more toward setting direction, developing capabilities, resolving difficult trade-offs and orchestrating human and AI resources.

For CEOs and other senior executives, AI therefore cannot be treated simply as another technology investment. It increasingly becomes a question of how the company should be organized and managed.

Organization DNA will become a differentiator

As AI technology becomes more accessible, the same technology can produce very different results in organizations with different DNA.

Some companies are accustomed to experimentation and can translate new ideas into action quickly. Others understand the need for change but remain trapped by existing processes, incentives or management habits. AI is likely to amplify these differences.

An adaptive organization may use AI as an opportunity to rethink how work is done. An overmanaged organization may simply add another layer of governance around it. A fragmented organization may generate many promising AI applications without ever turning them into a coherent enterprise capability.

This matters because the technological gap between companies may eventually narrow. Competitors can increasingly access similar models, tools and technical talent. But changing the way an organization makes decisions, shares knowledge, collaborates across boundaries, and learns from experience is much harder.

The lasting competitive advantage may therefore come less from possessing a particular AI technology and more from the ability to combine technology, people and the organization into a system that can continuously evolve. Such organizational capability is much harder to replicate.

A different question for CEOs

We are still at an early stage of this transformation, and some expectations will undoubtedly prove exaggerated.

Companies therefore face a familiar strategic challenge: They should neither chase every new development blindly, nor wait for the uncertainty to disappear.

If today’s organization were being designed from scratch with AI already available, would its processes, management layers and division of work look the same? For many companies, I believe they would not.

The first AI race has been largely about building intelligence. The next will be about deploying it — and, ultimately, learning how to organize around it. This may ultimately determine which companies merely become more efficient with AI and which ones are genuinely transformed by it.

This article was originally published in China Daily on September 30, 2026.