📰 📡 Latvia sees first use of AI in a political dispute
Minister of Agriculture Armands Krauze published a Facebook post stating that a premium artificial intelligence tool had assessed the origin of a letter sent by the Progresīvie parliamentary faction. According to the post, the AI suggested with “85–95% probability” that the information or draft materials used by the faction could have come from a well-known figure in Latvia’s timber sector. Krauze later submitted a request to the Prosecutor’s Office to examine potential breaches of transparency rules in the preparation of earlier government decisions related to timber-sector policy.
The Progresīvie letter asked for clarifications regarding a 2023–2024 report prepared by the Ministry of Agriculture on competitiveness issues in the timber industry. Krauze stated that similarities in phrasing and references prompted him to ask the AI tool for an assessment. The Prosecutor General candidate, Armīns Meisters, noted that when senior officials identify grounds for review, the prosecution may carry out an assessment; however, the launch of a criminal process is a last resort and is not automatic.
Context:
This episode marks the first public attempt in Latvia to use artificial intelligence for attributing authorship in a political dispute. Large language models do not have access to private correspondence, cannot determine document origin, and cannot provide forensic evidence; the “probability” figures they generate reflect narrative confidence, not statistical certainty. The case highlights the need for clearer guidance on the use of AI tools in public administration and political communication. Further developments will depend on the Prosecutor’s assessment and on whether additional rules emerge for the use of AI-assisted analysis in political processes.
Image: photos/photo_55@20-11-2025_13-49-50.jpg