What Is Policy Intelligence?

1. Why a New Capability Is Needed

Public policy has always dealt with uncertainty. What has changed is the speed, scale and interconnectedness of that uncertainty.

A technological development can create consequences simultaneously for labour markets, education, national competitiveness, regulation and public trust. A demographic shift can affect housing, healthcare, social policy and fiscal sustainability. Climate, artificial intelligence, biotechnology and new mobility systems increasingly cross the boundaries between traditional government portfolios.

Governments therefore face two related challenges.

The first is an information challenge. Decision-makers have access to more information than ever before, but greater information does not automatically produce greater understanding.

The second is a time challenge. By the time a conventional research cycle has identified a question, commissioned a study, collected evidence and produced a final report, the environment around the original question may already have changed.

Strategic foresight and anticipatory governance have consequently become more important across the public sector. OECD research has argued that governments need stronger capabilities to identify signals, explore plausible futures and connect anticipation more systematically with policy development and innovation.

Policy intelligence responds to this challenge by treating understanding the policy environment as a standing capability, rather than an activity activated only when a study is commissioned.

2. Defining Policy Intelligence

The Institute for Policy Intelligence defines policy intelligence as:

Three elements of the definition matter.

It is continuous

Policy intelligence does not begin when a request for a report arrives and end when the report is published.

Signals continue to emerge. Evidence changes. Policies create new effects. Public expectations evolve. Decisions therefore need feedback loops.

Policy intelligence continuously reconnects observation, analysis, policy design, implementation learning and decision support.

It is integrative

Government institutions frequently possess excellent research, data, foresight, consultation and evaluation capabilities. The challenge is that these capabilities may sit in different units and operate at different moments.

Policy intelligence connects them.

The Institute's institutional model combines four forms of intelligence:

It is decision-oriented

Research asks an important question: What do we know?

Policy intelligence continues: What is changing? What might happen next? What choices are available? What are their trade-offs? What should be tested? What does the decision-maker need to know now?

The output may still be a research report. But it might instead be an early-warning brief, scenario analysis, policy options paper, dashboard, experimental design, decision aid or implementation note.

The form follows the decision.

3. Policy Intelligence Is Not Another Name for Policy Research

Policy intelligence depends on rigorous policy research. It does not replace it.

But it is broader.

• Policy research develops evidence and understanding about a policy question.
• Policy analysis compares interventions, impacts, costs, trade-offs and implementation considerations.
• Strategic foresight systematically explores plausible futures rather than predicting one future.
• Horizon scanning identifies weak signals, emerging trends and potential disruptions.
• Data analytics extracts patterns and insight from quantitative information.
• Public insight introduces lived experience, expectations and perceptions.
• Policy experimentation tests assumptions before interventions are expanded.

Policy intelligence brings these capabilities into one operating logic organized around decision-making.

This is the distinction.

A government may possess excellent research teams, excellent data and excellent foresight capabilities and still lack a policy intelligence system if those capabilities are not connected to one another and to the rhythm of actual decisions.

4. The Policy Intelligence Cycle

The Institute operationalizes policy intelligence through the Policy Intelligence Cycle, an eight-movement method that moves from early signals to action and learning. The cycle is shown sequentially for clarity, but in practice it is continuous.

Then the cycle begins again.

A decision changes the environment. That change generates new evidence and new signals.

5. What Policy Intelligence Looks Like in Practice

Consider a government trying to understand the implications of rapidly advancing autonomous AI agents.

A conventional approach might begin by commissioning a study on the technology. Several months later, the study would describe the technology, international developments and potential regulatory issues.

A policy intelligence approach begins earlier.

• Sense: identify rapid advances in agent capabilities, investment, government use cases and regulatory signals.
• Understand: examine implications for public services, employment, cybersecurity, liability, procurement, privacy and accountability.
• Design: develop options ranging from guidance and procurement controls to pilots, regulatory frameworks or controlled experimentation.
• Simulate: explore failure scenarios, adoption trajectories, institutional impacts and unintended consequences.
• Test: pilot selected applications within defined boundaries.
• Evaluate: compare outcomes against performance, trust, risk and public-value measures.
• Advise: provide leadership with clear choices and implementation pathways.
• Communicate: explain selected policy directions appropriately to affected stakeholders.

The difference is not simply speed.

The difference is that research, foresight, design, experimentation and decision support operate as one system.

6. The Role of Artificial Intelligence

AI makes policy intelligence increasingly feasible at a scale that was previously difficult to sustain.

AI can support:

• continuous monitoring of large information environments
• evidence discovery and classification
• comparative policy mapping
• synthesis of large document collections
• identification of patterns and anomalies
• scenario generation
• simulation and stress-testing
• stakeholder and systems mapping
• rapid drafting of alternative policy options
• monitoring of policy outcomes

But capability should not be confused with authority.

AI can widen the field of what a policy team can examine. It can reduce analytical friction and reveal connections that may otherwise be missed.

It cannot determine what society should value, which trade-offs are acceptable or who should bear a particular risk.

7. From Research Capacity to Intelligence Capacity

The policy institution of the future will still need excellent researchers.

But it will also need the organizational ability to connect research with data, anticipation, design, experimentation, evaluation and decision-making continuously.

That changes the institutional question.

Instead of asking only: How can government produce better policy research?

The more important question becomes: How can government build a standing capability to sense earlier, understand faster, design better, test sooner and learn continuously?

That is the proposition behind policy intelligence.

It is not a report.

It is a capability.

Key Takeaway

Selected References

1. OECD, Towards Anticipatory Governance Guidelines for Public Sector Organisations (2025). OECD research emphasizes connecting foresight, experimentation and innovation more systematically to policy development.
2. OECD, Supporting Decision Making with Strategic Foresight (2023).