AI SAFEGUARDS · SHOCK ABSORPTION · CRISIS RESPONSE

Government Preparedness Index

A country-by-country comparison of observable public capacity to prevent avoidable AI harm, absorb economic and cyber shocks, and respond lawfully and effectively when disruption occurs.

This is not a probability, guarantee, endorsement, or prediction of government performance. It measures documented public capacity. Digital capability can enable faster public response, but it can also accelerate harmful AI deployment when safeguards and accountability are weak.

Countries: 195
Global mean: 53.9
Global median: 53.7
Highest observed score: Netherlands 87.1

OBSERVED PUBLIC CAPACITY

World preparedness map

Switch between the overall index, its three stages, and six factors. Dark red marks the least observed capacity. Red, orange, amber, and green mark progressively higher documented capacity on the same 0 to 100 scale used in every country profile.

Map capacity layer

100%

0100

Current country scores use solid dark red through red, orange, and amber to green on the 0–100 scale. Hover or focus for details, select a country to open its profile, and use the controls or drag after zooming. Greenland uses Denmark's profile data. Small-state dots use reviewed coordinates. The selected layer changes the map, not the published overall score.

0 comments · 0 votesOpen discussion

No comments yet. Start the discussion.

COUNTRY ANALYSIS

Capacity, bottlenecks, and confidence

Filter by region or weakest stage, then sort on any stage or factor. Confidence falls when a public source lacks a country observation and the model uses a disclosed income-group median.

Country nameRegionWeakest stageSort195 of 195 countries

Country	Overall	Weakest stage	Prevent	Absorb	Respond	AI safeguards	Digital state	Cyber	Social protection	Institutions	Crisis response	Confidence
Netherlands 🇳🇱	87.1Strong observed preparedness	Respond	86.3	92.6	82.9	80.9	95.4	99.2	85.9	83.9	81.0	86%
Norway 🇳🇴	86.5Strong observed preparedness	Prevent	82.5	91.5	87.0	76.0	93.2	97.0	85.9	88.0	85.0	86%
Denmark 🇩🇰	85.3Strong observed preparedness	Prevent	76.8	93.0	89.9	63.8	98.5	100.0	85.9	88.3	93.0	66%
Germany 🇩🇪	85.2Strong observed preparedness	Respond	83.2	91.9	81.4	76.8	93.8	97.9	85.9	81.1	82.0	86%
United Kingdom 🇬🇧	83.9Strong observed preparedness	Respond	81.4	93.0	78.7	72.7	95.8	100.0	85.9	76.1	84.0	86%
Finland 🇫🇮	83.6Strong observed preparedness	Prevent	75.8	93.0	85.8	63.8	95.7	100.0	85.9	87.7	82.0	66%
Estonia 🇪🇪	83.5Strong observed preparedness	Prevent	80.2	90.5	81.4	69.9	97.3	95.0	85.9	79.6	85.0	86%
Sweden 🇸🇪	83.0Strong observed preparedness	Prevent	74.9	92.6	85.3	63.8	93.3	99.3	85.9	84.9	86.0	66%
Iceland 🇮🇸	82.9Strong observed preparedness	Prevent	76.1	92.4	83.3	63.8	96.7	98.8	85.9	84.0	82.0	66%
Luxembourg 🇱🇺	82.3Strong observed preparedness	Prevent	71.6	92.3	88.3	63.8	84.7	98.7	85.9	87.9	89.0	66%
Ireland 🇮🇪	82.2Strong observed preparedness	Respond	80.7	88.4	78.3	74.2	91.4	90.9	85.9	83.0	69.0	86%
South Korea 🇰🇷	82.0Strong observed preparedness	Prevent	76.0	93.0	80.1	63.5	96.8	100.0	85.9	76.7	87.0	86%
Switzerland 🇨🇭	82.0Strong observed preparedness	Prevent	74.7	88.6	85.8	65.5	90.0	91.3	85.9	86.2	85.0	86%
France 🇫🇷	81.9Strong observed preparedness	Respond	80.0	92.5	74.9	75.5	87.4	99.0	85.9	72.9	79.0	86%
Japan 🇯🇵	81.3Strong observed preparedness	Prevent	70.8	91.8	86.7	57.1	93.5	97.6	85.9	85.1	90.0	86%
Spain 🇪🇸	81.3Strong observed preparedness	Respond	79.1	92.8	73.8	71.3	92.1	99.7	85.9	72.7	76.0	86%
Lithuania 🇱🇹	80.7Strong observed preparedness	Prevent	77.0	89.3	77.5	68.5	91.1	92.7	85.9	78.3	76.0	86%
Portugal 🇵🇹	80.7Strong observed preparedness	Prevent	75.1	92.9	77.1	69.7	84.2	99.9	85.9	74.1	83.0	86%
Australia 🇦🇺	80.4Strong observed preparedness	Prevent	72.9	91.1	81.0	59.1	95.8	96.2	85.9	83.5	76.0	86%
Slovenia 🇸🇮	79.9Substantial observed preparedness	Respond	76.4	91.2	74.2	69.6	87.6	96.5	85.9	74.3	74.0	86%

THREE STAGES · SIX FACTORS

A chain is constrained by its weakest stage

The overall result is a weighted geometric mean. Strong crisis response cannot fully cancel weak prevention, and a strong policy framework cannot fully cancel weak shock protection or institutional execution.

40%

Prevent avoidable AI harm

AI safeguards25%
Digital state15%

30%

Absorb economic and cyber shocks

Cyber resilience15%
Social protection15%

30%

Respond lawfully and effectively

Institutions20%
Crisis response10%

Sources: Global Index on Responsible AI 2026, UN E-Government Development Index 2024, ITU Global Cybersecurity Index 2024, World Bank Worldwide Governance Indicators, ILO World Social Protection Report 2024-26, and INFORM Risk 2026.

Responsible-AI, digital-government, cyber, and crisis-response gaps use disclosed income-group medians. Social protection uses the ILO 2023 income-group estimate for every country. Each fallback reduces confidence.

Read the complete method and formula →

FREQUENTLY ASKED QUESTIONS

Questions about this page

Short answers to common questions about the page, its evidence, and the limits of what DoomBench claims.

What does the Government Preparedness Index measure?

It compares observable public capacity to prevent avoidable AI harm, absorb economic and cyber shocks, and respond lawfully and effectively across 195 countries.

Is a high score an endorsement of a government?

No. The score is not an endorsement, a forecast, or proof that institutions will perform well in a future crisis. It summarizes documented capacity under a disclosed method.

Why use a geometric mean?

Preparedness depends on a chain of prevention, absorption, and response. A geometric mean limits how far exceptional strength in one stage can compensate for a serious weakness in another.

What happens when country data is missing?

The model uses a disclosed income-group median only when a source lacks a country observation. Every fallback lowers confidence, and incomplete source collection cannot replace the last validated dataset.