AMENINȚĂRILE FIMI ÎN PERSPECTIVA VIITOARELOR ALEGERI PARLAMENTARE DIN REPUBLICA MOLDOVA DIN 2025: O ABORDARE PROBABILISTICĂ BAZATĂ PE INFERENÞA BAYESIANÃÃRILE FIMI ÎN PERSPECTIVA VIITOARELOR ALEGERI PARLAMENTARE DIN REPUBLICA MOLDOVA DIN 2025: O ABOR

Foreign Information Manipulation and Interference (FIMI) has become an increasingly sophisticated tool for external actors seeking to influence democratic processes in Eastern Europe, particularly in Moldova. This study employs Bayesian inference to analyze the probability of distribution of the key FIMI-related threats in the context of the 2025 Moldavian parliamentary elections. By integrating the Open Source Intelligence (OSINT) data, governmental reports, and academic studies, five plausible scenarios were evaluated: (1) intensification of AI-driven disinformation campaigns, (2) exploitation of security fears to shape public perception, (3) financing and infiltration of pro-Russian political parties, (4) advanced disinformation campaigns aimed at undermining European integration, and (5) artificial mobilization of protests and public destabilization. The Bayesian model applied assigns likelihood ratios to each scenario based on the weight of evidence collected, revealing that AI-driven disinformation (30.5%) and security fear exploitation (21.8%) are the most probable threats. These findings highlight the urgent need for a multi-tiered information security strategy, incorporating advanced detection technologies, enhanced international cooperation, and strengthened media literacy programs. The study concludes that proactive monitoring and continuous scenario probability updates are crucial for mitigating the risks of FIMI in Moldova’s electoral processes.