As AI reduces the cost of prediction toward zero, the premium shifts decisively to judgment. And judgment, at the organizational level, requires a kind of organizational design that many enterprises have never built. The Manifesto for Enterprise Agility, released in early 2026 on the 25th anniversary of the original Agile Manifesto, is an attempt to define what that architecture looks like. It was developed through a landmark collaboration between PMI, the Agile Alliance, and the Agile Business Consortium, with input from accredited delivery partners and practitioner contributors. 

That timing is not coincidental. It is a deliberate signal that we have reached an inflection point. What began as a software delivery movement must now become an organizational operating system. Jim Highsmith, one of the original 2001 signatories, contributes a voice to this new document; a sign of continuity and evolution, not rupture.  

Ajay Agrawal's Prediction Machines put it precisely: cheap prediction makes judgment scarce. The manifesto is, at its core, a design guide for building organizations that can exercise it. 

The original Agile Manifesto was a direct response to the bureaucratic weight of plan-driven development; it gave teams permission to be human, adaptive, and focused on working outcomes rather than comprehensive documentation. Its impact was profound and lasting. While Scrum found its global stage, frameworks like Kanban and SAFe, and an entire consulting industry grew from those twelve lines. Practitioners working with large enterprises have documented cases where teams transitioned to shorter delivery cycles and began delivering value in days rather than months.

But here is the honest reckoning: the original manifesto was never designed for the enterprise system. It was designed specifically for software development teams. Its authors wrote it as a response to a software crisis, not an organizational one, and to their great credit they never claimed otherwise. The manifesto spoke of individuals and interactions, of working software, of customer collaboration. It said nothing about annual budgeting cycles, governance boards, talent strategy, portfolio prioritization, or cross-functional operating models, because those were not its concern.

The practitioner community did not wait for a formal manifesto to act. Many consulting firms had been advising large enterprises on operating model redesign, organizational restructuring, and enterprise-wide agility long before this document was published. Through their work with Fortune 500 companies, national governments, and regulated industries, they accumulated extensive evidence about what prevents Agile from scaling: governance structures misaligned with delivery cadence, funding models designed for capital projects rather than continuous iteration, and leadership behaviors that reward predictability over adaptability.

The Manifesto for Enterprise Agility does not necessarily introduce a new idea to the enterprise world. What it does offer, and this, in my view, is its genuine contribution, is the codification of that accumulated field knowledge into a shared, open framework that any organisation can draw upon, regardless of whether it has access to a consulting firm. It makes insights derived from hard experience available to all leaders and gives the broader leadership community a shared language for confronting a challenge that every organisation has faced, until now, in complete isolation. That challenge, as this article will argue, is the result of a single recurring error: organizations applied a team-level solution to a system-level problem, and that mismatch is precisely why so many are now structurally exposed to the pressures that follow.

“The Manifesto for Enterprise Agility does not necessarily introduce a new idea to the enterprise world. In my view, its genuine contribution, is the codification of that accumulated field knowledge into a shared, open framework that any organisation can draw upon”

The World We Are Now Managing In

Consider the decisions a Canadian CEO or board must now make simultaneously. A professional services firm headquartered in Toronto must decide whether to accelerate its US market presence at precisely the moment that trade relations between Canada and the United States have been thrown into their most volatile state in generations. Re-imposed US tariffs on Canadian goods have forced boards across manufacturing, energy, agriculture, and financial services to rethink supply chain architecture, pricing models, and customer contracts in real time.

These are not strategic planning questions that can be answered in an annual offsite. They require real-time sensing, rapid scenario modelling, and the authority to act; the defining capabilities of an enterprise-agile organization.

“Organizations that cannot adapt their structure as fast as AI reshapes their industry will find the productivity gap widening, not closing, regardless of which tools they deploy.”

Then there is the productivity crisis, which is Canada's most consequential long-term economic challenge¹ and the one most directly addressed by enterprise agility. Canada consistently underperforms OECD peers on business productivity and private-sector R&D investment². In knowledge-intensive industries, such as legal services, financial analysis, consulting, software development, and public-sector program delivery, AI tools are now performing tasks that defined entire job categories.  Organizations that cannot adapt their structure as fast as AI reshapes their industry will find the productivity gap widening, not closing, regardless of which tools they deploy.

Taken together, these forces create a Canadian operating environment in which the core question for senior leaders is no longer "What is our plan?" It is "How fast can we update our plan, and do we have the organizational design to act on the update?" What the Toronto firm and the knowledge-sector employer share is not a strategy problem. They share a structural one: each faces an environment that changes faster than their organizational architecture was designed to absorb. That is the systems problem this article turns to next.

This is where the intellectual contribution of the new manifesto becomes clear, and where systems thinking provides the essential frame for understanding why previous attempts at enterprise agility have underdelivered.

An organization is not a collection of teams. It is a system: a set of interconnected feedback loops, incentive structures, information flows, decision rights, and resource allocation mechanisms that together produce behavior. W. Edwards Deming observed that 94% of organizational failures are failures of the system, not of the individuals working within it. He was right in the manufacturing era. He is right in the AI era.³

When you optimize one component of a system without addressing the system's structure, you produce what systems thinkers call a local optimum, a team that is genuinely agile operating inside an organization that structurally prevents that agility from generating enterprise value.

Regulated industries make this dynamic particularly visible. In financial services and public administration, governance frameworks are traditionally built for compliance and auditability, not for adaptive decision-making.

I encountered this dynamic directly while working with product teams inside a large enterprise that had genuinely embraced Agile. Delivery teams were shipping features on a reliable quarterly cadence. But those features then waited an equal or greater period in a queue, held back by shared services, compliance checks, and legal review functions that operated on their own separate cycle. Value was being produced on one rhythm and released on another, and the gap between them was invisible to leadership because no single dashboard showed it.

This pattern is not uniquely Canadian. In engagements with large public sector organizations, sprint-based policy development cycles of roughly three months are achievable, and by government standards, genuinely fast. But development speed is only half the picture. Stakeholder alignment, successive governance approvals, and operational activation routinely add a year or more beyond the development timeline. Policies developed rigorously and on schedule are frequently never activated at all, not because the work was poor, but because the system surrounding it was not designed to receive it.

In both instances the diagnosis was the same: the team had been made agile. The system had not. We had built a local optimum: high-performing delivery inside an organizational design that could not receive what we produced.

The Enterprise Agility Manifesto addresses the system. Its nine principles, organized across leadership behavior, organization design, and execution, are not standalone practices, they are interdependent mechanisms. Governing with guardrails rather than gatekeepers only works if decision authority has genuinely moved to where value is created. Funding purpose rather than activity only works if leaders have real-time visibility into where value is being generated. Sensing early and acting with confidence only works if the data infrastructure beneath the organization is governed, trusted, and accessible.

“Enterprise agility is not the sum of agile departments. It is what happens when the systems connecting those departments, how decisions are made, how information flows, how resources are allocated; are designed to move together. Making each team faster in isolation does not produce an agile organization. Redesigning how the teams connect does.”

What the New Manifesto Gets Right for Senior Leaders

For executives, three dimensions of the manifesto deserve particular attention.

The governance model shifts from gatekeepers to guardrails. Gatekeepers made sense when only senior leaders had the information needed to make safe decisions. AI changes that: it can assess risk and model scenarios in minutes, putting the same information in everyone's hands. Guardrails define the boundaries within which teams can decide and act without waiting for approval, while keeping accountability intact. The result is not less governance. It is governance placed closer to the work.

The funding model shifts from fixed budgets to flexible investment. Annual budget cycles made sense when forecasting was expensive and locking in capital early was the prudent move. That logic no longer holds when scenario planning is available to any team at negligible cost. The real constraint today is not the ability to plan, it is that most organizations cannot reallocate resources once the plan changes. When a trade disruption or a competitor move rewrites the strategic picture mid-year, annual budgets leave leaders with the right judgment and no room to act on it.

The AI and data imperative is the most urgent, and the most misunderstood. Most organizations treat it as a technology problem. It is a judgment problem. You cannot deploy AI at enterprise scale without enterprise agility, and you cannot achieve enterprise agility without governed data. These are inseparable. When prediction is cheap, the constraint shifts to the quality of judgment that acts on it. Ungoverned data does not produce cheap prediction, it produces cheap noise, which in a fast-moving decision environment is more dangerous than no signal at all. AI fails not because the model is wrong, but because the organizational system was not designed to receive what it produces. That is the same diagnosis as enterprise agility failure. They are the same problem and require the same solution.

For Canadian leaders, the question is no longer whether to pursue enterprise agility. The trade pressures, the public sector transformation imperative, and the productivity gap have already decided that. The question is whether to build the organizational design that makes agility possible before the next disruption forces the issue. Those who do will not merely respond faster. They will be the ones whose judgment shapes what comes next.

“You cannot deploy AI at enterprise scale without enterprise agility, and you cannot achieve enterprise agility without governed data.”

---

¹Rogers, C. (2024, March 26). Time to break the glass: Fixing Canada’s productivity problem [Speech]. Bank of Canada. https://www.bankofcanada.ca/2024/03/speech-carolyn-rogers-march-26-2024/