Centre for Investigative Journalism

Over the past decade, disinformation has been the subject of intensive analysis, with researchers examining it from a multitude of perspectives. Among the chief architects of disinformation are governments, which often find themselves playing a double-edged role: both fueling the fire with propaganda and manipulative narratives while simultaneously crafting laws and regulations to combat disinformation. This article seeks to unpack a paradoxical dynamic where governments straddle the line between instigators and regulators of disinformation. Drawing insights from cross-country comparative studies, it delves into the interplay between the independence of public and state-controlled media and key factors like democratic quality, media freedom, and public integrity, in nations that have adopted anti-disinformation legislation. The findings unveil that such laws, which frequently curtail press freedom, are predominantly championed by authoritarian regimes or flawed democracies, wielding them as instruments of censorship. What sets alarm bells ringing, however, is the ripple effect of these practices, which are making inroads into countries known for their strong democratic foundations and well-established traditions of media freedom.

Language models are successfully applied to the problem of analysing crime descriptions from a police database with the purpose of prioritising suspects for an unsolved crime, given details of solved crimes. The frequency of terms in each description relates to the behaviour of the offender and this can be used to link crimes to a common offender. Language modelling uses Bayes' theorem and thus require a prior probability. Such a prior can be based on each offender's past propensity to offend, derived from historic data. Language modelling yields a probability of a document being relevant, which in this case is interpreted as the probability of a suspect being the culprit. Although the absolute value of the probability does not carry any direct applied implications, the study does show that the general likelihood of identification of the actual suspect does correspond to the relative values. Thus these probabilities can be used for more than just ranking suspects.