Smart Data Should Be Built as a Network — Not a Collection of Schemes
…and why smart data is like ice cream!
It was encouraging to see the UK Government publish Smart Data 2035: The UK’s Smart Data Strategy.
For those of us who have followed this space since the early days of the Data Portability Project (circa 2007) and the UK’s MiData initiative in 2012, this moment has been a long time coming. Progress has been slow, often fragmented—but now there is a clear marker in the ground.
A ten-year vision.
Multiple named schemes.
And a direct link to economic growth.
All of that matters.
But reading the strategy raises a more fundamental question:
Are we about to build twenty smart data schemes… when what we really need is one smart data network?
The Risk of Fragmentation
The strategy outlines six new schemes in the near term, and potentially twenty over time.
At one level, that sounds like progress.
But at another, it risks repeating a familiar pattern:
Multiple schemes
Multiple standards
Multiple user journeys
Multiple interpretations of “consent”
In other words:
Twenty ways to fragment the same underlying problem.
Because at a technical level, all smart data schemes are doing the same thing:
Moving data between endpoints
Verifying identity
Managing permissions
Logging and auditing exchanges
Which brings us to an unexpected but useful analogy.
Why Smart Data is Like Ice Cream
Smart data can come in many flavours:
Banking
Energy
Telecoms
Health
Retail
Each with its own nuances, regulations and use cases.
But underneath, the base ingredients are the same. At the base level they are all just 1’s and 0’s on silicon chips accessed by applications through an operating system.
Just like ice cream.
Different flavours. Same underlying construct.
Which raises an obvious question:
If the fundamentals are consistent… why are we designing entirely separate systems?
From Schemes to Infrastructure
Instead of building isolated schemes, we should be thinking in terms of Digital Public Infrastructure (DPI).
That means treating smart data like we already treat:
Payments networks
Telecommunications
Identity systems
Energy grids
These are not point-to-point solutions.
They are networks.
They work because:
Any participant can connect
Interactions are standardised
Trust is embedded in the system
No central party controls every interaction
And crucially:
They scale because they are designed as networks from the outset.
The Missing Piece: A Smart Data Network
What’s needed is not twenty schemes.
It’s a single, interoperable smart data network that can support all of them—and many more.
A network where:
Individuals, organisations and systems are all nodes
Data can flow any-to-any, not just within scheme boundaries
Interactions are governed by shared rules and standards
Every exchange is verifiable, auditable and secure
Think of it as:
Recorded delivery for personal data — at national scale.
Why This Matters Now: The Rise of Agentic AI
This shift becomes even more urgent when viewed through the lens of AI.
We are moving rapidly from:
Static interfaces
Manual data entry
One-off permissions
To:
Conversational interfaces
Continuous interactions
Autonomous, agent-driven behaviour
In this world:
AI agents will become the primary actors in data exchange—not humans.
They will:
Request data
Share data
Update data
Act on data
All on behalf of individuals.
That simply does not work in a fragmented, scheme-by-scheme model.
It requires a network.
Three Game-Changing Implications of a Network Model
1. Individuals Become First-Class Nodes
Instead of being passive subjects, individuals become active participants in the network.
With their own:
Secure identifiers
Data endpoints
Audit logs
Much like:
A bank account
A phone number
A postal address
This becomes:
A “data address” for the individual—owned and controlled by them.
With a full, verifiable history of:
What data moved
Where it went
Under what terms
A “bank statement for your data.”
2. Contracts Replace Consent
Today’s model relies heavily on consent.
We know its limitations:
Fatigue
Lack of understanding
Weak enforceability
A network model enables something stronger:
Standardised, machine-readable contracts governing every exchange.
Frameworks like IEEE P7012 (MyTerms) point the way forward.
They:
Encode permissions and expectations
Enable automation and enforcement
Provide clarity for both individuals and organisations
Put simply:
A contract can function as consent—but with far greater precision, transparency and accountability.
3. Data Flows Any-to-Any
Current models (e.g. Open Banking) are largely one-directional and tightly scoped.
A network model unlocks:
Data flowing to me, not just to third parties
Data flowing back to organisations (e.g. updates, corrections)
Data flowing between domains (finance, health, retail, etc.)
Data being continuously enriched and improved
All under individual control.
New Patterns That Emerge
Once you think in networks, entirely new use cases become possible.
Pattern 1: “Give Me My Data”
Individuals collect and curate their own data
Augment it using AI
Share it selectively
Pattern 2: “Blend and Share”
Multiple data sources combined
Enriched into a high-value dataset
Shared with advisors, providers or services
These are not theoretical.
They are already emerging through data intermediaries — a new class of infrastructure.
The Role of Data Intermediaries
Data intermediaries act as trusted connectors within the network.
But a critical distinction is emerging:
Who do they work for?
We are likely to see three models:
Individual-aligned (fiduciary)
Organisation-aligned
Attempted “neutral”
The most transformative is the first.
A fiduciary model — similar to doctors, lawyers or financial advisers—ensures:
Alignment of incentives
Duty of care
Trust by design
This will be essential in an AI-driven world.
What Happens If We Get This Wrong?
If we continue with fragmented, scheme-based thinking:
We recreate today’s complexity at scale
We entrench poor consent experiences
We limit the potential of AI
We weaken trust instead of strengthening it
In short:
We solve the wrong problem more efficiently.
What Happens If We Get This Right?
If we build a smart data network as digital public infrastructure:
Individuals gain real control and visibility
Businesses access richer, higher-quality data
Innovation accelerates across sectors
AI works with people, not around them
And critically:
Trust becomes a feature of the system—not an afterthought.
A Moment of Choice
The UK has a genuine opportunity to lead here.
But this is a fork in the road.
We can build:
Twenty separate schemes
Twenty consent flows
Twenty integration challenges
Twenty user experiences
Or we can build something far more enduring:
A national smart data network.
A Final Thought
We don’t need twenty flavours of smart data infrastructure.
We need one digital public infrastructure implementation that serves them all.
Because the long-term success of Smart Data in the UK will not be defined by how many schemes we launch…
…but by whether they can ultimately:
connect, interoperate, and evolve into something bigger than themselves.
Footnote and a Call to Action
If you are:
Designing smart data schemes
Regulating data exchange
Building platforms or AI systems
Then the question is NOT:
“How do we implement another scheme?”
But:
“Are we building towards a network—or away from one?”
We’re currently partnering with a small number of Organisations and Partners to explore these ideas through targeted Proofs of Concept. If you’re thinking seriously about the future of Smart Data, AI, and individual data control - we’d be interested in hearing from you.