Chapter 01 · Policy Synthesis · UK · EE · CA · CN · US

The GovernmentThat GovernsBy Design — 2030

A composite blueprint built from the strongest digital and AI strategies of five nations. Convenience without surveillance. Automation with accountability. One front door, no single point of failure.

Source States5
Capability Layers7
Target Horizon2030
Scroll to begin the story
The Synthesis

Each nation contributes its sharpest edge.

No single country has solved digital government. But each has built something the others lack. The composite takes the best — and deliberately leaves the worst behind.

01
ESTONIA

Digital Foundation

Cryptographic national ID, the once-only principle, and the X-Road exchange layer. Services online by default; paper as the rare exception.

02
UNITED KINGDOM

Delivery Discipline

A single GOV.UK front door, rigorous service-design standards, open-source-first procurement, and a central function that stops duplicated spend.

03
CANADA

Rights & Trust

Mandatory algorithmic impact assessments, a tiered directive on automated decisions, inclusive obligations, and a right to explanation.

04
CHINA

Scale & Integration

Super-app integration, deployment speed, and tight data-to-service feedback loops — adopted without the surveillance and scoring dimensions.

05
UNITED STATES

Innovation & Security

Cloud-scale infrastructure, a deep govtech vendor market, the NIST AI Risk framework, zero-trust security, and modular API-first buying.

The Composite
One login. One conversational front door. A federated data layer that shares only what's needed — every access logged and visible. Every automated decision carries a risk tier, a plain-language explanation, and a human appeal.

The Core Tension Resolved

China's speed and integration, married to Canada-style rights guardrails.

The Architecture

Estonia's federated data model — not a central honeypot of citizen records.

The Craft

UK service discipline over US security depth and elastic cloud scale.

The Promise

One front door without one central database. Automation you can audit.

Strategic Capabilities

Seven layers must exist for it to work.

Layers one through five are buildable with money and time. Layers six and seven are where strategies usually fail. Tap any layer to expand.

01

Identity & Trust

tap to expand
  • Sovereign digital identity — cryptographically secure, widely adopted, cross-service
  • Consent & access management citizens can see and control
  • A trust registry: verifiable credentials, e-signatures, reliable authentication
02

Data

tap to expand
  • Secure interoperability backbone (X-Road model) — federated, not centralised
  • Master data management: authoritative single-source registries
  • Data governance: classification, lineage, quality, ownership
  • Open data and APIs by default
03

AI & Intelligence

tap to expand
  • AI lifecycle platform: build, test, deploy, monitor, roll back
  • Model risk management aligned to NIST / Canada-style frameworks
  • Conversational and agentic delivery — AI that acts, not just answers
  • Explainability and audit tooling for every decision
04

Service Delivery

tap to expand
  • Single front door with consistent UX
  • Reusable components: payments, notifications, scheduling, documents
  • Omnichannel continuity — start online, finish in person, no data loss
  • Life-event orchestration, not departmental silos
05

Infrastructure & Security

tap to expand
  • Sovereign or trusted cloud with elastic scale
  • Zero-trust architecture and continuous monitoring
  • Cyber resilience: incident response, recovery, tested continuity
  • Observable platform engineering — uptime, performance, cost
06

Governance & Assurance

failure pointtap to expand
  • Central digital authority that sets standards and prevents duplication
  • Mandatory algorithmic accountability — assessments, registers, appeals
  • Procurement for modular, API-first, anti-lock-in buying
  • Privacy and rights by design, with independent oversight
07

People & Operating Model

most underfundedtap to expand
  • In-house digital and AI talent — an intelligent buyer and builder
  • Multidisciplinary teams: product, design, data, engineering, policy
  • Digital and AI literacy across workforce and leadership
  • Platforms funded as continuous products, not one-off projects
  • Change and adoption capability for citizens and staff
The Strategic Insight

The technology was never the hard part.

Layers one through five are buildable with budget and time. Layers six and seven are where strategies die — the systems work, but governance gaps erode trust and weak operating models leave governments unable to sustain what they built. The winners in 2030 treat institutions, not infrastructure, as the real project.

Next — How NSW gets there
CHAPTER 02New South Wales · The Road to 2030

From Strong StartTo CompositeGovernment

NSW already leads Australia on responsible AI and digital service delivery. This is the honest baseline — and the phased roadmap that closes the gap to the 2030 blueprint, layer by layer.

Public Servants400k+
Active Missions5
Target Year2030
The Baseline · Where NSW Stands

An honest read of today, against the seven layers.

NSW didn't start from zero. It has real assets — and real gaps. Each card maps a 2030 capability layer to what exists in NSW right now.

Identity & Trust

Emerging

Digital licences and the Service NSW wallet are mature and widely used. A statewide NSW Digital ID is legislated and rolling out, but not yet the universal cross-service backbone.

↳ Service NSW Wallet · Digital Driver Licence · Digital Photo Card · NSW Digital ID (in rollout)

Data

Fragmented

Strong pockets — the Digital Housing Pipeline and open data portals — but no whole-of-government federated exchange layer. Data still siloed across agencies.

↳ Data.NSW · Digital Housing Pipeline (3D dataset by 2027) · agency-level registries

AI & Intelligence

Leading

A genuine national strength. NSW ran responsible-AI governance before ChatGPT, and now has a dedicated Office for AI, the AI Assessment Framework, and a world-first agentic-AI guide.

↳ Office for AI · AI Assessment Framework (AIAF) · Agentic AI Guide · NSWEduChat

Service Delivery

Leading

Service NSW is a national exemplar — one trusted front door across hundreds of transactions. The next leap is life-event orchestration and a conversational/agentic layer.

↳ Service NSW (app + centres) · 1,200+ transactions · Easy Read Hub

Infrastructure & Security

Maturing

Cyber NSW and cloud policy provide a base, plus resilience systems like EMMACS. Zero-trust and full sovereign-cloud elasticity are still in progress across agencies.

↳ Cyber Security NSW · GovDC / cloud policy · EMMACS emergency coordination

Governance & Assurance

Leading

Among the strongest layers. Mandatory AIAF, tiered residual-risk review, Test & Buy procurement, and a central Digital NSW authority already setting standards.

↳ Digital NSW (DCS) · AIAF compliance plans · Test & Buy Innovation

People & Operating Model

Building

Digital Skills is a named mission and the Office for AI is driving literacy. But in-house talent depth and product-style continuous funding remain the hardest nut — as everywhere.

↳ Digital Skills mission · Office for AI literacy uplift · Performance & Wellbeing Framework
The Verdict
NSW is ahead on the layers that usually fail — governance and AI assurance — and behind on the plumbing: a federated data exchange and a universal identity spine. The roadmap inverts the usual order: build the foundations under the strengths already standing.

Strengths

3

AI · Service Delivery · Governance

In Progress

3

Identity · Infrastructure · People

Critical Gap

1

Federated data layer

Horizon

2030

Composite outcome

The Roadmap · Today → 2030

Three phases. Foundations first, intelligence on top.

Each move is tagged to the capability layer it advances, connecting straight back to the blueprint. Sequence matters: data and identity must exist before agentic delivery becomes safe.

2026–27
Phase 1 · Foundations
Lay the plumbing under the strengths.
L1 ID

Scale NSW Digital ID from rollout to default — one verified login across Service NSW and agency services, consent-managed.

L2 DATA

Stand up a federated data-exchange backbone (X-Road model). Start with high-value flows: housing, transport, emergencies. Federated, not a central honeypot.

L2 DATA

Establish master registries — authoritative single sources for person, business, address — with the "tell us once" principle.

L5 SEC

Adopt zero-trust as standard across agencies via Cyber Security NSW; begin sovereign-cloud consolidation.

2027–28
Phase 2 · Integration
Turn services into one continuous experience.
L4 SVC

Re-architect Service NSW around life events — "new baby", "start a business", "after a disaster" — bundling across departments instead of by agency.

L4 SVC

Build shared reusable components: payments, notifications, scheduling, document handling — consumed by every agency.

L3 AI

Deploy an AI lifecycle platform — versioning, monitoring, drift testing — extending the AIAF from policy into operational tooling.

L7 PPL

Shift to product funding: fund platforms as continuously-funded products, not one-off projects; deepen in-house engineering.

2029–30
Phase 3 · Intelligence
Safe agentic government — convenience without surveillance.
L3 AI

Launch a conversational / agentic front door that acts — pre-filling from held data, completing transactions — built on the Agentic AI Guide.

L6 GOV

Make every automated decision explainable and appealable — published risk tier, plain-language reasons, human appeal route, transparency register.

L2 DATA

Citizen-visible access logs — every data touch viewable by the person it concerns. Trust as a feature, not a promise.

L7 PPL

Whole-of-service AI literacy reaching frontline and leadership, so adoption sticks and humans are freed for higher-value work.

Why NSW Can Win This

It already holds the hardest cards.

Most governments fail at layers six and seven — governance and operating model. NSW leads on exactly those. The work ahead is foundational, not cultural: build the federated data spine and universal identity, then let the intelligence layer the state is already good at do its job. The 2030 blueprint isn't a leap for NSW. It's the next three budgets.

Next — How it gets funded
CHAPTER 03The Investment Portfolio & Operating Model

Capital Up Front.Efficiency Out.Opex Freed.

How NSW funds the journey: a central enablement engine that invests capital to retire run-cost, with autonomous decision-rights and control devolved to agencies. Portfolio investment management for every digital decision in government.

Annual ICT Spend~$3.6B
Spent Just Running~80%
Demand That's Legacy~50%
The Problem · Why Funding Is Broken

The budget is trapped in keeping the lights on.

Government IT funding is structurally stuck. The vast majority of spend goes to running and maintaining what already exists — much of it ageing legacy — leaving almost nothing for transformation. Every year the run-cost grows, and the squeeze tightens.

Today — trapped budget$3.6B

RUN / MAINTAIN 80%CHANGE 20%

Target — 2030 rebalanced$3.6B

RUN 60%CHANGE 40%
~80%

of government IT budgets are consumed just running existing systems — and roughly half of all NSW digital investment demand over recent years went to modernising legacy, without clearly cutting the run-rate it creates.

The trap: legacy generates technical debt that compounds. Spend more to maintain it, and there's even less for change next year. Breaking the cycle needs capital deployed deliberately to permanently lower run-cost.

The Funding Thesis

Capitalise to free operating budget.

Rather than asking agencies for more recurrent funding, the model deploys capital up front to build shared platforms — which retire duplicated legacy and permanently lower run-cost, freeing existing opex to be redirected to services.

STEP 01

Deploy capital

A central fund invests capital up front in shared foundations — identity, data exchange, reusable components.

STEP 02

Retire legacy

Shared platforms replace duplicated agency systems. Legacy is decommissioned, not just wrapped.

STEP 03

Lower run-cost

Maintenance and licence run-rate falls permanently as the legacy estate shrinks.

STEP 04

Recycle the saving

Freed opex is redirected — part to services, part recycled into the fund for the next wave.

This is the proven NSW pattern: the Digital Restart Fund began at $100M and grew to $1.6B, explicitly created to fix a funding model CIOs considered broken — combining new capital with savings redirected from agencies.

Value Delivery Over Time

The J-curve: invest first, harvest after.

Portfolio value follows a deliberate shape. Capital goes in during the foundation years with little visible saving. As platforms land and legacy retires, the run-cost curve bends down and cumulative benefit compounds — crossing into net-positive and accelerating to 2030.

Portfolio cash & benefit profile · 2026 → 2030

ILLUSTRATIVE — shape of value, not budgeted figures
break-even 2026 2027 2028 2029 2030 value $ net +ve
Capital invested (per year, declining)
Recurring benefit released (compounding)
Cumulative net value (the J-curve)
The "Never Stops" Problems

Answering what business areas fear most.

Some costs feel permanent — cyber never stops, legacy is never solved. The portfolio doesn't pretend to end them; it changes their trajectory, converting open-ended escalation into a managed, declining run-rate.

Cyber Security

"Cyber never stops."

True — the threat is permanent and rising. You can never stop spending on it.

Correct — so the goal isn't to stop, it's to spend less per unit of protection. Consolidating to a zero-trust shared platform replaces dozens of duplicated agency defences with one continuously-funded capability. The threat stays; the fragmented, escalating cost does not.

Legacy Technology

"Legacy is never solved."

We've spent years modernising and the legacy estate is still here.

Because past spend modernised without measuring run-cost reduction. This model funds decommissioning as the deliverable — retirement, not wrapping — and tracks the falling maintenance bill as the success metric, not just the new system delivered.

Operating Budget

"We can't afford new opex."

There's no recurrent budget headroom for another platform to maintain.

That's the point — this is capital-funded, not opex-funded. It's designed to reduce your recurrent bill by retiring what you currently maintain, freeing existing opex rather than asking for more.

Agency Autonomy

"Central control slows us down."

A central fund means we lose control and wait in a queue.

No — the centre enables; agencies decide. You hold delegated authority to draw down against the portfolio within guardrails, choosing what to adopt and when. Central platforms, autonomous agency decisions.

The Operating Model

Central enablement, autonomous agencies.

The hardest balance in government technology: economies of scale without central bottlenecks. The model splits roles cleanly — the centre builds and assures the commons; agencies hold the decision-rights and control over their own adoption and outcomes.

The Centre Enables

Build · Assure · Steward
  • Funds the portfolio — holds capital and recycles realised savings into the next wave
  • Builds shared platforms — identity, data exchange, reusable components, AI lifecycle tooling
  • Sets standards & guardrails — security, assurance (AIAF), interoperability, accessibility
  • Assures the spend — stage-gates, benefits tracking, transparent portfolio reporting
  • Prevents duplication — buy-once, reuse-many across all clusters

Agencies Decide

Choose · Control · Own
  • Hold delegated authority — autonomous draw-down against the portfolio within guardrails
  • Choose what to adopt — and the sequence that fits their service priorities
  • Own the outcomes — accountable for the benefits and service results they commit to
  • Control their roadmap — no central queue gating day-to-day delivery
  • Report benefits realised — feeding the recycle loop that funds the next agency
Portfolio Investment Management

One discipline for every digital decision.

Best-practice portfolio management — applied not to one programme but to all government digital investment. Decisions are made as a managed portfolio: prioritised against drivers, staged, measured, and rebalanced continuously.

P1

Align to drivers

Every investment maps to core government drivers: protect core services, manage finances, anticipate future needs, grow the economy.

P2

Stage-gate funding

Capital released in tranches against proven progress — fund the next stage only when the last delivered. Fail fast, fund what works.

P3

Benefits as the metric

Success is measured in run-cost retired and outcomes delivered — not project completion. Benefits are tracked to the ledger.

P4

Reuse before build

Buy-once, reuse-many. New build is the exception, justified only when the shared commons genuinely can't serve the need.

P5

Rebalance continuously

The portfolio is reviewed and rebalanced each cycle — stopping underperformers, doubling down on what compounds value.

P6

Recycle the gains

Realised savings flow back into the fund, making the portfolio increasingly self-sustaining over time.

The Bottom Line

Fund it once. Free it forever.

The 2030 government isn't paid for with a bigger budget — it's paid for by unlocking the budget already trapped in legacy and duplication. Capital up front, deployed through a managed portfolio, retires the run-cost that strangles every agency. The centre builds the commons; agencies stay in control. Cyber and legacy never stop — but their cost curve finally bends down.

Back to the start