Artificial intelligence has advanced at an extraordinary pace over the past several years. Generative AI, large language models, and increasingly autonomous AI agents are giving organizations capabilities that would have seemed almost unimaginable not that long ago.
But technological advances don’t automatically translate into economic value. Throughout the history of computing, major technological innovations have required complementary changes in business processes, organizational structures, management practices, and workforce skills before their full benefits could be realized. AI is unlikely to be an exception.
This distinction between technological capabilities and organizational deployment is central to the argument made by Princeton computer science professor Arvind Narayanan and PhD candidate Sayash Kapoor in their influential 2025 essay, “AI as a Normal Technology.”
It was also one of the strongest recurring topics at the 2026 MIT Sloan CIO Symposium, whose overall theme was “Human-AI Synergy: Redefining Leadership for a Transformative Future.”
Taken together, these two perspectives suggest that the central challenge facing enterprises is shifting from developing and acquiring increasingly powerful AI technologies to transforming their organizations so they can use those technologies effectively.
From Technological Progress to Organizational Deployment
Artificial intelligence was founded as an academic disciplines in the summer of 1956 at a Dartmouth College workshop where researchers proposed the development of machines capable of performing tasks requiring human-like intelligence.
The early AI pioneers were remarkably optimistic about how quickly their objectives could be achieved. But after decades of unfulfilled promises and exaggerated expectations, AI experienced periods of reduced interest and funding, which became known as AI winters.
AI was successfully reborn in the 1990s with the emergence of data-centric approaches that combined sophisticated algorithms, increasingly powerful computers, and very large amounts of data. One prominent milestone occurred in 1997, when IBM’s Deep Blue supercomputer defeated world chess champion Garry Kasparov. Progress accelerated in the 21st century with big data analytics, deep learning neural networks, generative AI, and, more recently, agentic AI.
As The Economist observed in a September 2025 article, opinions about AI range from utopian visions of extraordinary economic growth and scientific breakthroughs to dystopian predictions of widespread job losses and potentially catastrophic consequences if and when AI achieves a kind of superintelligence.
But, it’s very difficult to predict how AI will evolve over the coming decades. As Narayanan and Kapoor explain, there is a critical distinction between advances in AI technologies — increasingly powerful models and ever larger computing infrastructures — and the much slower process through which those capabilities are deployed across organizations and the economy.
They argue that, — as has been true of previous general-purpose technologies including electricity, computers, and the Internet, — “diffusion is limited by the speed of human, organizational, and institutional change.” Their economic and social impact depended not simply on technological breakthroughs but on the complementary innovations required to make effective use of them.
The same is likely to be true of AI. “To view AI as normal is not to understate its impact — even transformative, general-purpose technologies such as electricity and the internet are ‘normal’ in our conception,” wrote Narayanan and Kapoor.
Their argument is not that AI will have only a modest impact. Rather, its impact will depend on how effectively people, organizations, and institutions learn to use it. Impressive advances in AI capabilities don’t necessarily translate into equally rapid improvements in enterprise productivity. Organizations must redesign work processes, develop new skills, establish governance mechanisms, integrate AI with existing systems, and adapt longstanding management practices.
From AI Adoption to Organizational Transformation
The distinction between adoption and transformation was one of the dominant themes of the 2026 MIT Sloan CIO Symposium. A recurring Symposium theme was that AI adoption depends less on adding intelligent technologies than on redesigning organizations and business processes to support them.
Now in its 23rd year, the Symposium brought together technology and business leaders to discuss the challenges and opportunities associated with the emerging age of AI. Its overall theme was “Human-AI Synergy: Redefining Leadership for a Transformative Future.”
This year’s Symposium included 20 sessions. For the first time, advances in AI made it practical to produce a comprehensive report examining the collective body of discussions, session summaries, planning documents, historical materials, participant feedback, and supporting research.
As Symposium Executive Chair Allan Tate explained, “AI made it possible to compare perspectives, identify recurring themes, and uncover relationships that would have been extraordinarily difficult to recognize through manual review alone.”
“Technology accelerated synthesis, but understanding remained fundamentally human,” he wrote.
This process represented a form of collective sensemaking — a new way of synthesizing knowledge across a professional community by leveraging technology to integrate multiple perspectives and develop a deeper understanding of complex issues.
Collective sensemaking illustrates the Symposium’s central theme of human-AI synergy. AI enabled a new form of analysis, while human judgment remained essential for interpreting the findings and understanding their implications.
Five Dimensions of Organizational Transformation
The report identified five recurring themes that, taken together, provide a useful framework for how enterprises should approach organizational transformation in the age of AI.
1. Leadership Has Become the Primary Technology Challenge
Perhaps the strongest consensus across the Symposium was that AI transformation is fundamentally a leadership challenge rather than simply a technical one. Technology expertise remains essential, but it’s no longer sufficient.
In earlier generations of enterprise computing, technology leaders were often primarily responsible for acquiring, developing, and operating information systems. Today, they must also help lead the transformation of the organizations those systems support.
The modern CIO must serve as a bridge between technology and the business, helping executive colleagues understand both the opportunities and the organizational changes required to realize them. As AI becomes increasingly integrated into just about all enterprise operations, CIOs must work closely with CEOs and other business leaders to translate technological capabilities into changes in business strategy, organizational design, and management practices.
The challenge is no longer simply how to deploy AI technologies, but how to lead organizations through the changes that AI makes possible.
2. AI Is Driving Organizational Redesign, Not Simply Automation
Many organizations initially approach AI as a way to automate existing tasks and improve operational efficiency. Symposium participants repeatedly distinguished between automating existing work and redesigning work itself. The greatest benefits of AI may come not from doing existing work more efficiently, but from learning how to organize work differently.
Automating a task can make an existing process faster or less expensive. But transforming a process requires reconsidering how the work should be performed in the first place. Such changes may require redefining roles, responsibilities, organizational boundaries, and management practices. They require rethinking longstanding assumptions about how organizations operate.
“Technology enables change. Organizations determine whether that change creates value.”
3. Governance Is Becoming a Strategic Capability
Traditional approaches to technology governance have often emphasized compliance, risk reduction, and control. But the Symposium presented governance as something more: a strategic capability that can enable organizations to innovate more confidently and deploy AI more broadly.
Enterprise practitioners described a variety of approaches, including AI councils, architectural standards, enterprise marketplaces, approval processes, monitoring systems, and clear business ownership.
Effective governance helps organizations establish the conditions under which AI can be safely and productively used. Without such mechanisms, enterprises may struggle to move beyond isolated experiments. Individual teams may develop promising AI applications, but organizations may lack the confidence, accountability, and operational controls required to deploy them at scale. Properly designed, governance can become an important enabler of innovation.
4. Enterprise Architecture Is Reemerging as a Strategic Asset
A somewhat unexpected finding was the renewed importance of enterprise architecture. Organizations often focus on the capabilities of AI models while underestimating the technical foundations required to integrate those capabilities into enterprise operations.
These foundations include reliable enterprise data, identity management, application integration, shared platforms, reusable components, and the ability to monitor how systems behave. Without them, even highly capable AI systems may remain confined to isolated applications.
Solid foundations are particularly important as organizations move toward complex autonomous agents capable of accessing information, interacting with applications, and performing complex tasks.
5. Human Capabilities Become More Important, Not Less
Public discussions about AI frequently assume that increasingly capable machines will inevitably reduce the importance of human expertise. The Symposium presented almost the opposite conclusion. Even highly technical discussions repeatedly returned to the importance of organizational trust, executive communication, workforce readiness, continuous learning, and cultural adaptation.
As AI systems become capable of performing more complex tasks, the nature of human work will undoubtedly change. But organizations will continue to need people who can define problems, exercise judgment, evaluate AI-generated recommendations, manage exceptions, and decide when automated systems should or should not act. The challenge is not merely to teach employees how to use AI tools, but to redesign jobs and develop the human capabilities needed for effective collaboration with increasingly intelligent systems.
The Symposium’s theme — Human-AI Synergy — has proved to be much more than a slogan. It reflects a broad consensus that the successful deployment of AI will depend on combining technological capabilities with human judgment, organizational trust, and effective leadership.
Conclusion: From Technology Leadership to Organizational Leadership
“Perhaps the Symposium’s most enduring insight is that AI is changing the nature of enterprise leadership more profoundly than it is changing enterprise technology,” noted the Symposium report.
The extraordinary advances in AI capabilities are making increasingly sophisticated technologies available to organizations of all sizes. But access to those capabilities will not, by itself, determine which organizations succeed. The greater challenge will be learning how to redesign work, develop new organizational capabilities, establish effective governance, modernize enterprise architectures, and help employees adapt to new ways of working.
The diffusion of transformative technologies is constrained by the pace of human, organizational, and institutional change. AI may continue advancing at an extraordinary pace. But the future of enterprise AI will depend as much on our ability to transform organizations as on our ability to develop more intelligent machines.
AI disclosure: I wrote the initial draft of this blog and used ChatGPT for editorial assistance. I remain responsible for the final content.