Digital Statecraft in the Age of AI: Rethinking How Markets Are Coordinated
A keynote at the Digital Statecraft Academy Residency week 2026
We are living through a profound shift in how economic systems are coordinated. Much of the current discourse focuses on AI, platforms, and increasingly, the idea of decentralisation. But beneath all of this sits a more fundamental question—one that is often overlooked:
What is the unit of coordination in our economy today?
To answer that, we need to step back and look at how economic society has evolved.
For most of history, markets came before the state. Early exchange was local, human, and relational. Over time, institutions emerged to scale that exchange—rules, norms, and systems that allowed strangers to transact safely. The state then evolved as a stabilising force, governing markets, enforcing contracts, and enabling coordination at scale.
Each technological shift has reshaped this relationship. Today, we are in the midst of another such transition—driven not just by AI, but by something more foundational: data.
From Firms to Flows: The Changing Nature of Coordination
At the meso level—the level of markets and firms—we are already seeing a structural transformation.
Firms are not disappearing, but they are becoming thinner. What was once internal—IT, HR, finance, workflows—is increasingly externalised into cloud-based services and platforms. Coordination itself is being outsourced.
Large firms now operate as coordination nodes, sitting within a web of external services rather than vertically integrated entities. At the same time, smaller firms are becoming more powerful—more global, more efficient, and less constrained by size, largely because they can plug into these same coordination infrastructures.
This shift has increased output and efficiency. But it has also created a deeper problem.
The coordination layer of the economy now sits outside traditional institutional boundaries.
And this is where the macro-level tension begins.
A Crisis of Governance
If firms are thinner, more distributed, and more dependent on global coordination systems, then what exactly is the state governing?
Today, individuals, firms, and platforms operate globally, while governance remains largely local. The result is fragmentation. Jurisdictions blur. Taxation struggles to align with where value is actually created.
Much of the current policy response focuses on AI sovereignty—control over compute, models, and infrastructure. But this, I would argue, misses the point.
Because AI is not just a technology. It is coordination intelligence.
It acts on data. It accelerates decision-making. And crucially, it operates at a level far below where our institutions are currently designed to intervene.
Which brings us back to the original question:
Where does coordination actually begin?
The Shift to the Nano Layer
The answer is increasingly clear: coordination now begins at the nano level.
At this level, individuals and their data form the atomic units of the economy. Data is generated continuously through behaviour. AI systems act directly on these signals in real time. And value creation begins before firms or markets even enter the picture.
This is a profound shift.
It means that the foundation of economic activity is no longer transactions between firms, but interactions at the level of individuals and their data.
And yet, our institutional frameworks have not caught up.
The Misunderstanding of Data
There is a common tendency to treat data as if it were a natural resource—like oil or land—something that exists independently and can simply be extracted, owned, or redistributed.
But this is fundamentally incorrect.
Data is not “found.” It is created—intentionally, and for a purpose. And because of that, it is already institutionalised at the point of creation. It is embedded within the rules, norms, and systems of the organisation that generated it.
This creates a critical constraint.
Data is locked into its original purpose. It cannot easily flow across contexts. It cannot be recombined freely. And as a result, it cannot function as a general-purpose economic resource.
Why Decentralisation Falls Short
In response to this, many have proposed decentralisation—returning data to individuals, redistributing control, or building new technological infrastructures.
But decentralisation, in itself, does not solve the problem.
It shifts control, but it does not redefine the underlying institutional structures. It often assumes that technology can determine those structures, replacing one system logic with another. And it does little to address the fact that data remains bound to its original purpose.
Without addressing that, decentralisation risks creating fragmentation rather than coordination.
Re-Institutionalising Data
What is needed instead is a more fundamental shift: the re-institutionalisation of data.
In economics, institutionalisation is the process by which behaviours, rules, and norms become stable, patterned, and taken for granted within a system.
Data today is already institutionalised—but narrowly, and incorrectly.
To unlock its potential, we must re-embed it into new institutional forms that:
Separate data from its original purpose
Define clear rights, roles, and permissible uses
Enable reuse across contexts
Allow data to flow within and across markets
This is not simply a technical challenge. It is an institutional one.
From Data to Markets
When data is re-institutionalised, something important happens.
It becomes usable.
It becomes tradable.
It becomes a resource that can participate in economic activity beyond its point of origin.
This enables individuals to participate meaningfully in markets. It allows firms to access richer and more legitimate data flows. And it creates the conditions for entirely new forms of exchange to emerge.
In other words, it enables the transition from nano-level activity to micro-level markets.
Designing for a Coordination Economy
If we take this seriously, then the task ahead is not to build better platforms or more powerful AI systems.
It is to design institutional architectures for data.
This includes solving practical problems that we already encounter every day—data silos, repeated verification, lack of portability. Today, the same information must be submitted and verified repeatedly across institutions, because trust does not travel with the data.
The goal, therefore, is not centralisation or decentralisation, but portable trust.
This requires new design principles:
separating custody, orchestration, and agency
ensuring systems act only on instruction, not control
creating shared objects (such as credentials or “badges”) that carry rules and meaning across contexts
allowing markets to form asynchronously, without pre-integration
These are not abstract ideas. They are concrete steps toward making data function as an economic resource.
From Coordination to Statecraft
Ultimately, this is not just about markets. It is about governance.
When data is properly re-institutionalised:
micro-level interactions become simpler
meso-level markets become aligned
macro-level systems become more stable
Data becomes anchored, governable, and jurisdictionally meaningful—much like labour or capital.
This is the foundation of what we might call digital statecraft.
Making the Invisible, Valuable
Perhaps the most powerful implication of this shift is what it makes possible.
When data is re-institutionalised, it allows us to make the invisible visible—and therefore valuable.
Capabilities, credentials, activities, and experiences that were previously unrecognised can now participate in markets. Trust becomes portable. Inclusion becomes scalable. And capital can flow to places it previously could not reach.
The Real Question
So the question is not whether we should decentralise data.
The question is:
How do we design institutions that allow data to function as a true economic resource?
Because only then can we move from fragmented systems to coordinated ones.
Only then can markets and states operate effectively in an AI-driven world.
And only then can we fully realise the potential of the coordination economy.