Introduction:
Digital news ecosystems shape how crime, cyber threats, terrorism, espionage, and public security incidents become visible to decision makers. This study develops an AI structured crime and security intelligence framework for measuring visibility concentration, geographic asymmetry, significance stratification, and source specialisation in global security reporting.
Methods:
An original corpus of 182,207 crime and security related news records was assembled from global media providers using web scraping, RSS feeds, and application programming interfaces. After preprocessing, deduplication, country harmonisation, significance normalisation, and rating validation, the final dataset contained 164,268 unique URL event assignments, 25 event categories, 994 source domains, 140,533 resolved country assignments, and 164,147 valid severity ratings. Concentration metrics, chi square tests, nonparametric comparisons, Jensen Shannon divergence, and correspondence analysis were used to analyse the structure of AI mediated security visibility.
Results:
The results show a highly unequal security information environment. Civilian Murder Event and Government Data Dump of Citizens and Employees account for 53.18% of cleaned assignments. The United States represents 39.09% of resolved country labelled records, while the country distribution shows an extreme Gini coefficient of 0.970. Event category and significance class are strongly associated, χ² = 156,273.32, V = 0.399, p < 0.001. Severity ratings also differ substantially across event categories, H = 124,135.95, ε² = 0.758, p < 0.001. Correspondence analysis shows that the first two dimensions explain 74.62% of country category association inertia.
Conclusions:
This study shows how AI structured news analytics can convert global security reporting into measurable computational intelligence. The framework supports open source intelligence, security informatics, computational media analysis, and AI assisted situational awareness.