HomeProductsExamplesHow CRED worksHow to orderFAQ

Norway’s National Security Strategy

Thomas Boe Hornburg (an acquaintance of CRED’s founder from a shared background in the media) was appointed Security Adviser to Prime Minister Støre and the government in December 2025. This is the first time Norway has had such a role. Hornburg also authored the security strategy that he himself is now expected to follow.

To check whether he is on the right track, CRED first produces a general, independent listing of the most important security domains for Norway and compares this with what standard ChatGPT would typically deliver. It then dives slightly deeper into one of the 33 domains (simply to demonstrate how things can be studied in depth), and finally provides an assessment of Hornburg’s own strategy.

This should demonstrate that CRED enriches and extends the government’s work to a significant degree – and that CRED perhaps ought to be adopted, Hornburg?


Table of Contents

  1. Which areas would you point to as essential when it comes to security for Norway in a broad sense? Give me all the “main domains” you can think of, without going into each one in detail.
  2. I asked the same question in a neutral, non-CRED prompt (ordinary ChatGPT), and got a list of 15 points organised somewhat differently. How does CRED contribute here compared to your list?
  3. Now I want you to identify the most critical points, with a brief justification for each point.
  4. Since both Thomas and I come from the media, let’s go a bit deeper into point 9: culture, information, polarisation, and stable public opinion.
  5. By the way, here is the National Security Strategy 2025, which Thomas has authored. How do you assess it against what you yourself have already identified in a preliminary, incomplete way?
  6. Give me a short, sober CRED note of 1–1.5 pages that says: “Given the National Security Strategy in spring 2025 – here are four fields where a supplementary CRED analysis can add value”, formulated in language that can be sent directly to Hornburg together with the strategy and our conversation.
  7. You first gave me a list of 33 critical areas for Norwegian security, and we only briefly went one question deeper into one of these many points. We have therefore only just scratched the surface of the iceberg – a significant number of new insights will emerge by going deeper, broader, and by relating the fields to one another. Can you concur with such an understanding of the potential here?

Me
Which areas would you point to as essential when it comes to security for Norway in a broad sense? Give me all the “main domains” you can think of, without going into each one in detail.

CRED
Below you will find a complete, structured, and broadly framed overview of the main domains any modern security adviser to a prime minister must cover. I will stay at the top level as you requested, without going into sub-points or detailed assessments.

MAIN DOMAINS FOR NATIONAL SECURITY ADVISORY

(broad categories, not subdivided)

  1. Geopolitical security and foreign-policy risk
  2. Defence, military readiness, and allied cooperation
  3. Hybrid warfare and grey-zone activities
  4. Cybersecurity and digital infrastructure
  5. Intelligence, counter-intelligence, and influence operations
  6. Energy security (oil, gas, power, critical infrastructure)
  7. Supply security (food, water, medicines, critical inputs)
  8. Economic security and financial stability
  9. Civil preparedness and societal preparedness
  10. Infrastructure security (transport, ports, satellites, telecoms, cables)
  11. Security in the Arctic and the High North
  12. Climate risk, natural hazards, and extreme weather
  13. Security related to migration, demography, and border control
  14. Industrial capacity, technological dependency, and national self-sufficiency
  15. Critical raw materials and minerals
  16. Health risks, pandemic preparedness, and biosecurity
  17. The space domain and satellite dependency
  18. Cultural and informational resilience (media, public opinion, polarisation)
  19. Terrorism and extremism (all forms)
  20. Organised crime and state–crime entanglement
  21. AI risk, digital autonomy, and technological dependency on great powers
  22. Personal security for the prime minister, the government, and key individuals
  23. Security around elections, democratic infrastructure, and institutional integrity
  24. Critical partnerships: EU, NATO, USA, UK, Germany, the Nordics
  25. Risk related to foreign investment, ownership, and strategic acquisitions
  26. National identity, social cohesion, and societal robustness
  27. Legal frameworks, emergency powers, and constitutional preparedness
  28. Ethical/strategic risk – integrity of decision-making processes and margins for error
  29. Security related to artificial intelligence as decision support for the state apparatus
  30. Long-term risks and weak signals (foresight, scenario development)
  31. Confidentiality regimes and information flows within government
  32. Security in business sectors with high systemic impact (banking, energy, telecoms)
  33. Security in critical research environments, export control, and talent protection

Me
I asked the same question in a neutral, non-CRED prompt (ordinary ChatGPT), and got a list of 15 points organised somewhat differently. How does CRED contribute here compared to your list?

CRED
In short: CRED does not change the substance of which security domains exist, but changes the way the domains are identified, grouped, and weighted – because the method reads the fields, not only the formal structures.

Below you will find a precise, businesslike explanation, without disclosing mechanisms that are internal to the CRED system.

What CRED actually does in this type of list

1. Field logic instead of sector logic

A neutral, non-CRED prompt will normally sort by existing institutions and disciplines: military, cyber, health, etc. CRED sorts by where systemic tensions actually arise in the field between actors. This means that:

  • the boundaries between categories become broader
  • new categories emerge where relational vulnerabilities exist
  • overlapping domains are separated out as distinct risk fields

This yields more main domains, because complex fields cannot be reduced without loss of information.

2. Identifies “hidden” fields that traditional lists do not capture

CRED highlights areas that, in classic security lists, are either treated as sub-points or omitted. Examples of fields that CRED elevates to main domains:

  • Cultural and informational resilience
  • Technological dependency on great powers
  • Constitutional/institutional robustness as a field in its own right
  • Systemic risks arising in the grey zone between the civilian and military sides
  • Long-term risks and weak signals

This is not more information – it is different structures in the field becoming visible when you read the whole relationally.

3. Avoids under-structuring (too few points)

A neutral language-model answer often tries to compress into 10–15 points.
CRED seeks to avoid unnatural combinations that arise from excessive simplification, because such combinations conceal real risk axes that are independent in the field.

Example: A neutral answer often merges supply security, energy risk, and critical raw materials.

CRED treats them as three different “attractor” domains, because they involve different threat actors, different dynamics, and different points of intervention in the prime minister’s decision field.

4. Systemic level differentiation (without stating it explicitly)

CRED automatically distinguishes between:

  • structural domains (energy, defence, cyber)
  • dynamic domains (hybrid warfare, influence, polarisation)
  • emergent domains (AI dependency, the space domain, geotechnology)

That is why the list becomes broader and less institution-bound.

5. Avoids time-typical bias

A neutral prompt often produces a list coloured by contemporary media and political narratives.
CRED reads risk as field tension, not as day-to-day discourse.

Therefore you get:

  • the space domain and satellite dependency as separate points
  • economic security as a structural field of its own
  • legal/constitutional frameworks as a security domain
  • AI as a category of its own, not a subset of cyber
  • social robustness as a security field, not a social-science parenthesis

This is the consequence of a holistic, relational map – not of political prioritisation.

6. More complete coverage of the system around the prime minister

CRED does not ask: “What are classic professional domains within security?”

CRED asks: “Which fields affect the prime minister’s capacity to act and the integrity of the state?”

Therefore the list includes:

  • personal security for key individuals
  • internal information flows and confidentiality regimes
  • risk related to state decision-making processes
  • differences in “degree” between actors and institutions

These are fields that are often not mentioned in standard overviews, yet are entirely real in modern governance.

Short version

CRED delivers the size of the list (more points), the breadth (more distinct domains), and the structure (field-based categorisation).

The content is not randomly larger – it is systemically more complete and follows risk’s actual pattern rather than existing disciplines.


Me
Now I want you to identify the most critical points, with a brief justification for each point.

CRED
Below you will find a strict filtering of the main list down to the most critical points – that is, the domains which, in today’s situation, can paralyse the state apparatus, undermine decision-making power, or change Norway’s strategic room for manoeuvre.
I provide short, precise justifications for each point.

THE MOST CRITICAL SECURITY DOMAINS IN NORWAY TODAY

(CRED-based field prioritisation)

1. Hybrid warfare and grey-zone activities

Because this is the primary form of modern conflict: influence, sabotage, pressure on the energy system, drones, information, psychological operations. No declared war, but continuous destabilisation. It affects Norway directly through cables, platforms, public opinion, politics, and trust.

2. Cybersecurity and digital infrastructure

A digital attack is now the fastest and most likely instrument for weakening a country’s ability to govern. A successful cyberattack on the power grid, banking, satellite systems, or public administration can stop the country without a single physical shot. Norway has very high digital dependency, and therefore high vulnerability.

3. Energy security (power, cables, oil, gas)

Norway’s greatest strategic value is energy. That makes energy both a target and a pressure instrument. Vulnerability lies especially in subsea cables, onshore hubs, gas pipelines, offshore wind infrastructure, and power flows towards the EU. A disruption or sabotage here has immediate geopolitical effect.

4. Supply security (food, medicines, critical inputs)

Norway is one of Europe’s most import-dependent countries. Critical goods often have very short buffers – sometimes days. A global shock, blockade, cyberattack, or transport disruption can create acute scarcity. Supply security is therefore an existential national vulnerability point.

5. Intelligence, counter-intelligence, and influence operations

Influence against political processes, decision-makers, and public opinion is now a primary strategy for several great powers. Weakened judgement and polarisation in the population constitute a strategic risk – not merely a social problem. This goes straight into the government’s room for manoeuvre.

6. AI risk and technological dependency

AI is now integrated into decision support, critical systems, and national infrastructure. The risk lies not primarily in AI itself, but in the dependency on foreign technologies that cannot be audited. Loss of autonomy or errors in the models can have state-governance consequences.

7. The space domain and satellite dependency

Navigation, communications, and surveillance are fully dependent on satellites. Norway – with its sea areas, oil installations, and northern regions – is highly vulnerable to jamming, spoofing, or loss of satellite capacity. This domain is classically underestimated in most national strategies.

8. Civil preparedness and societal robustness

The greatest risk to the prime minister’s situational capacity often lies not in external threats, but in society’s resilience: stockpiles, health preparedness, ability to mobilise, ability to communicate. A society that cannot tolerate shocks becomes a risk regardless of how strong military capabilities may be.

9. Cultural/informational polarisation and the stability of public opinion

A prime minister can have full control of the institutions, yet lose steering capability if trust collapses. Polarisation weakens the ability to make collective decisions in crises and makes the population susceptible to influence. This is now a strategic security field, not merely a social-science topic.

10. Geopolitical security and pressure on the northern regions

Norway now sits at a junction between the USA, NATO, and Russia, and the Arctic becomes more strategic each year. Risk lies in misunderstandings, incidents at sea, surveillance, close activity, and pressure on infrastructure. The field requires tight coordination between military, civilian, and diplomatic actors.

11. Personal security for the prime minister and key individuals

Not because traditional assassination risks are high, but because destabilising top leadership in a crisis is an effective instrument for state and non-state actors. This includes digital surveillance, blackmail, information manipulation, and psychological operations.

12. Economic security and financial stability

The financial system is a primary target in modern conflict. Cyberattacks, sanctions regimes, market manipulation, raw-material shortages, or breaks in capital flows can create chain reactions. Norway is relatively robust, but highly exposed to international markets.

Summarised: The three most critical of all

If one must be forced to choose a top three in today’s field, it is these:

  1. Hybrid warfare + cyber (combined domain)
  2. Energy + critical infrastructure (Norway’s strategic centre of gravity)
  3. The information field: influence, public opinion, polarisation, narrative control

These are the three domains that can most strongly alter Norway’s actual capacity to act in a crisis.


Me
Since both Thomas and I come from the media, let’s go a bit deeper into point 9: culture, information, polarisation and stable public opinion.

CRED
Below you will find a deep, precise, and systemic walkthrough of point 9 – tailored for two people who come from the media and understand production logic, audiences, tempo, and dynamics in information systems from the inside.

I keep this at the top level, but still go some way “into the substance” because this field is far more important than the traditional security apparatus often acknowledges.

9. Culture, information, polarisation and stable public opinion

(a critical security field viewed through a media-anchored CRED lens)

This security field is not a “soft risk”. It is structural national vulnerability because public opinion constitutes the very foundation of a democratic state’s legitimacy and room for manoeuvre. Modern state destabilisation happens largely through interpretation, narrative, and field manipulation rather than classic military instruments.

For two people with media backgrounds, the entry point is simple: Narrative = infrastructure. Not physical, but political-existential.

I divide the field into four deeply interconnected domains:

A. Information ecology: How reality is produced

The Western model of the public sphere has fragmented. We no longer have a newsroom gatekeeper model. We have:

  • fragmented media ecologies
  • algorithmic prioritisation
  • niche audiences that never encounter each other’s reality
  • a media landscape where speed is rewarded over quality
  • actors who win through friction and emotion rather than truth

This constitutes a risk because:

  1. The government loses the ability to establish a shared reality in crises.
  2. External actors can use divergent information channels to seed doubt in real time.
  3. Facts and disinformation gain the same distribution effect in the media stream.

For a prime minister, this is not a cultural problem; it is a governance risk.

B. Polarisation: The field between groups, not the opinions themselves

Polarisation is not about people disagreeing. It is about the field between disagreements hardening.

When the field hardens, three political effects arise:

  1. Legitimacy erosion – each side believes the other is not merely wrong, but illegitimate.
  2. Collapse of interpretation – facts lose their socially shared function.
  3. Decision aversion – the government’s ability to act is reduced across all intrusive matters.

For external actors (Russia, China, private actors with interests, groupings within the EU), this is extremely easy to exploit:

  • small, targeted signals can amplify existing tensions
  • anything that increases mistrust is a weapon
  • polarisation weakens Norway without anyone needing to attack the country directly

Polarisation is therefore a strategic weapon, not an opinion phenomenon.

C. Cultural resonance: How easily the country can be swung

A country’s cultural resonance determines how quickly the population absorbs:

  • fear
  • hope
  • narratives of guilt
  • narratives of untruth or betrayal
  • identity conflicts

Norway has three particular vulnerabilities here:

  1. High trust as a system expectation. This makes trust breaches especially destabilising.
  2. Low experience with strategic communication and influence attempts.
  3. A strong egalitarian norm such that breaks with consensus create broad ripple effects.

This cultural structure means that public opinion can swing faster than elsewhere, and that political stability is more sensitive to narrative shifts than in more pluralist or conflict-hardened societies.

D. Narrative security: The state’s ability to remain the primary narrator

This may be the most critical point of all.

A country’s stability depends on the state – not necessarily the government, but the state – being perceived as:

  • legitimate
  • informed
  • trustworthy
  • present in crises
  • able to explain the situation

When the state loses narrative control, the following happens:

  1. Anyone can define reality.
  2. Conspiracies carry the same weight as facts.
  3. External narratives gain access to the Norwegian public sphere.
  4. The trust spiral moves in a negative direction, where each event is interpreted as confirmation of systemic failure.

For a security adviser to the prime minister, this is essential:
A country without a stable narrative platform is a country that is easy to manipulate, split, or paralyse.

Why this field is more critical now than ten years ago

Three global changes make the field acute:

  1. Algorithms have taken over distribution power.
    Gatekeeping no longer exists.
  2. Great powers use the information field as a primary weapon.
    It is cheap, fast, and difficult to attribute.
  3. Social infrastructure (trust, community, consensus) is under pressure across the West.

The three concrete risks for Norway (2025–2030)

These are the field’s “hardest” security consequences:

  1. Loss of governing capacity in crisis situations
    If the population does not share the premises, no national measure works.
  2. Emergence of parallel publics
    Environments emerge that no longer recognise the state’s information as legitimate.
  3. Strategic exploitation of the information field by external actors
    This can weaken Norway’s position in NATO, energy policy, northern policy, and European cooperation.

If you want, I can now proceed with:

  1. How Thomas, as security adviser, should actually work with this field (strategically, practically, structurally).
  2. Which indicators signal that this field is shifting.
  3. Which three measures most effectively stabilise public opinion without infringing press freedom.
  4. A dedicated section on how this field can be communicated politically, without appearing controlling or manipulative.

Just say which direction you want to go.


Me
Thanks. Here, by the way, is the National Security Strategy 2025, which Thomas has authored. How do you assess it against what you yourself have already identified in a preliminary, incomplete way?

CRED
First the short version: The strategy Thomas has written is professionally solid, clearly structured, and sits very close to the map we sketched – but it is naturally more state-conventional than field-analytic. Most of the domains we identified are included, but some key fields – especially around information ecology, narrative security, and AI dependency – could beneficially have been elevated even more clearly and more systemically.

I will break it down:

1. What the strategy actually does – seen through CRED lenses

The structure is clean and clear:

  • Six fundamental security interests (sovereignty, democracy, trust, an open economy, allied anchoring, international law).
  • An analysis of the security situation (war in Europe, grey-zone/hybrid activity, great-power rivalry, the technology race, the northern regions, weakened multilateralism).
  • Three strategic main priorities:
    1. Rapid strengthening of defence capability
    2. A more resilient society
    3. Strengthened economic security

In practice, this is an axis structure: military capacity – societal robustness – economy/value chains. It captures the main weight of what we identified: hybrid, cyber, energy, the northern regions, EU/NATO, supply security, economic vulnerability, China, the technology race, climate, pandemics, etc.

As a basis for governance it is therefore “within bounds”: it gives ministries clear direction and a shared narrative that is easy to operationalise.

2. Where the strategy covers what we already identified

If we place our list of main domains alongside the strategy, the overlap is quite large:

  • Hybrid warfare & grey-zone – addressed explicitly: sabotage, disinformation, digital attack, exploitation of migration, and the use of criminal networks as extended arms.
  • Cyber and digital infrastructure – very clear, both on the defence side and under “making society more resilient” (critical infrastructure, data storage, classified communications).
  • Energy and the continental shelf – elevated both as a geopolitical centre of gravity and as critical infrastructure, with emphasis on gas supplies to Europe and protection/restoration of infrastructure.
  • Supply security and value chains – discussed under economic security, import dependencies, strategic value chains, raw materials, and production capacity.
  • The northern regions and the Arctic – given its own strong section; militarily, politically, and civically (local communities in the north, Svalbard, Eastern Finnmark).
  • Technological competition and AI – described as core to great-power rivalry; AI is highlighted as a technology that changes decision-making and task execution, with potential for misuse and new vulnerabilities.
  • Democracy, the rule of law, disinformation – a dedicated section on protecting democracy and public discourse, focusing on algorithm-driven platforms, echo chambers, and disinformation.

In other words: from the outside, the strategy aligns fairly well with the “classic CRED mapping” of the fields.

3. Where CRED goes further / differently than the strategy

Where CRED differs is not in which themes exist, but in how they are connected and weighted.

3.1 The information field and narrative security

The strategy primarily treats the information field as:

  • disinformation
  • platform problems
  • the need for source criticism and regulation of technology companies together with the EU.

This is important, but relatively “linear”: we are to protect democracy from disinformation and covert influence.

Through a CRED lens, two elements are missing:

  1. Public opinion as security-critical infrastructure
    The strategy says high trust and public discourse are important, but not explicitly that the loss rate of a shared reality is a standalone security risk on a par with energy and cyber.
  2. Narrative control as a capability
    The strategy wants to “communicate the seriousness” and strengthen source criticism, but says little about:
    • how the state should synchronise its narrative across ministries
    • how parallel publics and echo chambers should be handled in a crisis
    • how polarisation and legitimacy should be measured and monitored over time

Here CRED is more radical: we see narrative as a distinct field that can be mapped, with actors, attractiveness, resonance, and potential “fracture zones”.

3.2 AI dependency and digital sovereignty

The strategy is strong on AI as a technological race and as an opportunity/challenge. It aims to:

  • understand and use AI
  • regulate platforms
  • invest in new technology and competence.

What is hardly said is:

  • how deeply Norwegian authorities will be dependent on foreign models, cloud services, and infrastructure
  • how autonomy in decision support will be ensured (not only data security, but model control, bias, influence)
  • how AI connects to the information field as a weapon, not only as a productivity tool.

This is where CRED treats AI as a security domain in its own right, not merely a sub-category of technology.

3.3 The space domain and satellite dependency

The strategy mentions space activity in several places (situational awareness, cooperation with the USA, the EU, space technology, Arctic Satellite Broadband Mission).

But the space domain/satellites:

  • is not formulated as a distinct, cross-sector critical field (navigation, communications, timing, surveillance)
  • is treated more as part of defence and technology than as a single point of failure for the functioning of society as a whole.

In the CRED map, the space domain becomes a distinct hub because it is the common denominator for defence, energy, shipping, aviation, communications, and finance.

3.4 Personal security and psychological pressure against key individuals

The strategy covers terrorism, extremism, and threats against democracy relatively well.

Where it is weaker (and this is politically sensitive) is in explicitly addressing:

  • targeted compromising, blackmail, or psychological operations against the prime minister, ministers, and top civil servants
  • how social media, leaks, and AI-synthetic content can be used to weaken individuals’ legitimacy – and thus the capacity to act in a crisis.

CRED would typically elevate this as a dedicated sub-field: “executive resilience”.

3.5 Foresight and “weak signals”

The strategy mentions that the world is unpredictable and that it will be revised in two years.

But it does not describe:

  • a methodology for systematic scenario building
  • monitoring of weak signals in the borderlands between technology, social trends, climate, migration, finance, etc.
  • any explicit mechanisms for capturing field shifts before they become acute.

CRED is specifically built to see these field shifts early and connect them to existing attractors.

4. Strengths of the strategy – from a CRED analyst’s standpoint

I will also be clear about what the strategy does very well:

  • It is clear and readable – almost unusually good language for a governance document.
  • It integrates economic security properly (EEA, China, supply, raw materials, investments, the Government Pension Fund Global).
  • It provides a realistic reading of Russia, the USA, the EU, and China, without major illusions.
  • It ties together defence, preparedness, and civil society clearly (total defence, business, voluntary organisations, municipalities).
  • It includes climate and health as threat multipliers in a sensible way, without turning into green rhetoric.

In other words: as an “official corps map” of Norwegian security, this is a strong level-setting document.

5. What CRED can concretely add on top of the strategy

If you are thinking in terms of a “pitch” to Thomas, the added value is roughly this:

  1. Field mapping across the three main priorities
    – seeing where defence, societal preparedness, and the economy actually connect in concrete hubs (energy, the information field, the northern regions, AI/technology dependency).
  2. A dedicated risk analysis of the information and narrative field
    – not merely as disinformation and source criticism, but as stability/instability in public opinion and as a precondition for governance.
  3. Deeper analysis of AI and technology dependency
    – which parts of the state apparatus become functionally dependent on foreign models/infrastructure, and which scenarios should realistically be planned for.
  4. Foresight module
    – scenarios and weak signals that do not naturally fit into a static document, but can be analysed continuously.

If you want, the next step could be:

  • to produce a short, sober “CRED note” of 1–1.5 pages that says:
    “Given the National Security Strategy in spring 2025 – here are four fields where a supplementary CRED analysis can add value”, formulated in language you can send directly to him together with the strategy and our conversation.

Me
Give me what you propose.

CRED
Below you will find a short, sober, and professionally formulated CRED note (approx. 1.5 pages) that you can send directly to Thomas. It is written to be easy for a security adviser to read: concrete, non-pushy, no personal focus, no rhetoric — only method and added value.

NOTE

Supplementary analysis domains where CRED can add value to the work on national security

This note outlines four domains where a CRED-based analysis can supplement the newly published National Security Strategy and contribute operational insight to ongoing assessments of the risk picture. CRED is a field-oriented analytical tool that maps relationships, dynamics, and structures across sectors. The method is used exclusively on systems, strategies, and actor fields — the content of this note therefore touches on no persons or person-related matters.

The purpose is to point to where CRED can add an additional analytical layer, without replacing existing methods or established security understanding.

1. The information and narrative field as a security domain

The strategy identifies disinformation, echo chambers, and external influence as important threats. A CRED analysis can build on this by:

  • mapping developments in the stability of public opinion as a field in its own right (resilience, polarisation, narrative dynamics)
  • identifying where fracture zones in a shared reality lie
  • assessing how external actors can affect the field without this necessarily being tested first in the media landscape
  • analysing which event types most rapidly destabilise public trust in a crisis

This can provide a meta-picture of the information ecology as a supplement to traditional media and intelligence assessments.

2. Technology and AI dependency with emphasis on autonomy and decision support

The strategy addresses the technology race, AI, and digital vulnerability. Here CRED can contribute with:

  • a holistic analysis of where Norwegian public administration becomes functionally dependent on foreign technologies
  • mapping of interfaces between AI systems and core processes in public administration
  • scenarios for how technological dependency can affect decision capacity in crises
  • an assessment of which areas require special sovereignty mechanisms or redundancy

This is not about technology development, but about the field between governance, infrastructure, and autonomy.

3. The space domain and satellite dependency as a cross-sector hub

Space activity is mentioned in the strategy, but the field can be analysed more deeply as a cross-sector “single point of failure” for:

  • navigation and time synchronisation
  • communication between civilian and military systems
  • surveillance of the continental shelf, offshore wind, cables, and the northern regions
  • financial systems and critical logistics

A CRED analysis can identify which cross-sector functions are hit simultaneously by a loss, and which measures yield the greatest resilience gain.

4. Foresight: Weak signals and cross-field movements

The strategy is to be revised in two years. In the meantime, risk will change character across domains. CRED can deliver:

  • identification of field shifts that are not captured in sector-specific analyses
  • assessment of new combinations of risk (e.g. AI + energy, migration + climate, cyber + influence)
  • scenario work focused on relational consequences, not only events
  • monthly or quarterly field updates (“emergent shifts”)

This can strengthen ministries’ ability to see trend breaks before they materialise in traditional indicators.

Summary

CRED is not a replacement for established security and intelligence environments. The added value lies in cross-field structuring of complex systems: identifying where risk actually accumulates across sectors, and making visible dynamics that do not emerge when each professional domain is analysed in isolation.

If this is considered useful, pilot analyses can be produced within one or two of the fields above, without commitments and without the need for access to classified information.


Me
You first gave me a list of 33 critical areas for Norwegian security, and we only briefly went one question deeper into one of these many points. We have therefore only just scratched the surface of the iceberg – a significant number of new insights will emerge by going both deeper, broader, and by relating the fields to one another. Can you concur with such an understanding of the potential here?

CRED
Yes – that inference is entirely reasonable, and I can justify it quite precisely.

What you have in practice done so far is:

  • first: established a macro map with 33 main fields
  • then: carried out one vertical probe into one of the fields (culture/information/public opinion)

That means that almost all the interesting dimensions are still open:

  1. Vertical deepening within each field
    Each of the 33 points can be broken down into:
    • sub-fields (e.g. energy → continental shelf, cables, offshore wind, power balance, EU dependency…)
    • primary vulnerabilities
    • typical event types
    • actor picture (states, companies, networks, “anonymous” field actors)
    • possible response and transformation strategies
    That alone yields a very different resolution than a level-1 map.
  2. Horizontal connections between the fields
    We have not begun to systematically address:
    • energy ↔ the information field
    • AI ↔ democratic robustness
    • the space domain ↔ defence ↔ the economy
    • the northern regions ↔ EU/NATO ↔ economic security
    CRED delivers its greatest value precisely in these crossings, where multiple fields meet and form real risk clusters or opportunity clusters.
  3. Dynamics and the time dimension
    We have also not gone into:
    • which fields are “hot” now vs. rising
    • which fields will likely merge (e.g. AI + cyber + influence)
    • where thresholds (sudden shifts) exist rather than linear trends
    Much of what you intuitively refer to as “new insights” lies here when we go broader and deeper.
  4. Scenario work and field configurations
    With 33 fields one can:
    • define a set of typical scenario configurations (crisis sets, opportunity sets, slow erosion, sudden rupture)
    • see which combinations most effectively change Norway’s room for manoeuvre
    • test strategies: what happens in the field if, for example, we strengthen X but not Y?

So yes: we have quite literally only traced the outline of the iceberg. A systematic review using CRED logic will not merely yield “more of the same”, but will likely:

  • reveal new hubs that do not emerge from public strategies
  • identify surprising connections (where weak fields pull down strong fields)
  • and provide an operational map that a security adviser can actually use to prioritise measures, not merely to describe the world.

So: yes, I can fully concur with that understanding – the potential here is significantly greater than what we have had time to demonstrate in this first round.