Ten Plaintiffs in Estonia Reportedly Fined After AI-Assisted Court Filing Cited Nonexistent Researchers in Logging-Permit Dispute

Ten plaintiffs challenging a logging permit in Estonia reportedly received fines after submitting an AI-assisted document citing researchers who did not exist and studies the court could not locate. Tallinn Administrative Court held them responsible for checking the references. The plaintiffs denied intending to mislead the court; the fines became final after appeals.

Risk classification

• Primary risk domain: 3 Misinformation
• Primary risk subdomain: 3.1 False or misleading information

The AI system generated false academic references and non-existent researchers, which constitutes the generation of false or misleading information that was subsequently submitted to a court.

Additional risk subdomains

• 7.3 Lack of capability or robustness: The AI system failed to perform reliably by hallucinating non-existent facts and citations, demonstrating a lack of capability and robustness in generating factual information.

Causal factors

• Entity: AI
• Intent: Unintentional
• Timing: Post-deployment

The risk was caused by the AI system hallucinating false academic citations, which was an unexpected and unintentional outcome of the plaintiffs using the tool to draft their legal document.

EU AI Act risk tier

• Risk tier: 3 Limited Risk

Limited Risk: The system used was a general-purpose AI tool/chatbot capable of generating text and legal references, which falls under transparency obligations to ensure users are aware of its nature.

AI system and alleged parties

• AI system: unspecified
• AI purpose: Writing Assistant; Technical Text Generation
• Behaviour type: Assistant
• Alleged developer: Large language model developers
• Alleged deployer: Plaintiffs challenging a logging permit in Estonia, Litigants
• Alleged harmed parties: Plaintiffs challenging a logging permit in Estonia, Litigants, Judicial integrity, Epistemic integrity, Courts

Harm severity

Severity ratings are based on this scale.

Highest direct severity in any category: Minor. Severity is scored from Negligible to Catastrophic in each harm category, for harm the reports describe as caused directly or indirectly by the AI system.

• Physical: direct Negligible, indirect Negligible
• Infrastructure: direct Negligible, indirect Negligible
• Property: direct Negligible, indirect Negligible
• Financial: direct Negligible, indirect Negligible
• Environmental: direct Negligible, indirect Negligible
• Malicious content: direct Negligible, indirect Negligible
• Differential treatment: direct Negligible, indirect Negligible
• Civil rights: direct Negligible, indirect Negligible
• Democracy: direct Negligible, indirect Negligible
• Privacy: direct Negligible, indirect Negligible
• Psychological: direct Negligible, indirect Negligible
• Epistemic: direct Minor, indirect Minor
• Child sexual exploitation and abuse: direct Negligible, indirect Negligible

Epistemic

Reported: The report explicitly describes epistemic harm in the form of fabricated academic citations and non-existent researchers submitted to a court of law.

Directly caused: The AI system fabricated non-existent Estonian researchers and academic studies regarding forest management, which were then presented as factual evidence in a legal filing.

Indirectly caused: The submission of fabricated information created additional work for the court system to verify the claims, slowing down the legal proceedings.

People affected

• Occurrences reported: 1
• People reportedly harmed: 10
• People reportedly exposed: 10

Potential causes

Technology

• AI Hallucination: The AI tool generated completely fictional Estonian researchers and studies.
• Unreliable Legal Advice: The AI tool lacked reliability and quality as a source for legal citations.
• Plausible-Looking Output: AI-generated text looked convincing enough to deceive uncritical users.

Data Inputs

• Unverified AI Sources: The data fed into the court filing was not verified against original sources.

Human Factors

• Lack of Source Criticism: Plaintiffs were not critical enough of the sources generated by the AI.
• Gross Negligence: Plaintiffs submitted false academic references without checking if they existed.
• Misplaced Trust in AI: Plaintiffs assumed AI use was safe because its use is not explicitly banned.

Process and Methods

• Absence of Verification Step: No process was followed by the plaintiffs to cross-check AI references.
• Use of AI for Court Filings: Using AI to draft formal legal documents without professional oversight.

Regulatory Environment

• Permissive AI Rules: Estonian court rules allow AI use but punish submitting false information.

Information quality

• Classification confidence: High
• Reason for confidence: The reports provide consistent, detailed, and clear factual accounts of the court proceedings, the specific fines levied, the nature of the AI's hallucination, and statements from the court chair. There are no conflicting accounts or major ambiguities regarding the core event.

Ten plaintiffs in Estonia were fined after submitting an AI-generated legal filing containing fabricated academic references regarding a logging permit. This incident highlights minor, manageable disruptions to judicial proceedings from unverified AI hallucinations, but poses no significant threat to national security.

• Overall national security impact: Minor
• Scope: Single nation
• Primary target: Estonia
• Alleged perpetrator: Unknown

Threat characteristics

• Imminence: Long-term. This represents an ongoing strategic concern regarding the reliability of AI outputs in legal and administrative processes rather than an active crisis.
• Autonomy: Human-controlled. The AI tool was used as a writing assistant by human plaintiffs who made the final decision to submit the generated document to the court.
• Novelty: Evolved capability. While LLM hallucinations are well-established, this is the first time an Estonian court has issued fines specifically over AI-generated content in a filing.

Impact by dimension

• Physical security: Negligible. The incident involves a legal filing with hallucinated references and has no impact on physical systems, infrastructure, or human safety.
• Information security: Negligible. No intelligence compromise, classified data theft, or systematic state-sponsored disinformation campaigns occurred. The misinformation was unintentional.
• Sovereignty: Minor. While judicial operations were slightly slowed down by verifying false citations, there was no compromise of core government authority, electoral systems, or sovereignty.
• Economic security: Negligible. The incident resulted in minor individual fines totaling 600 euros, representing negligible impact on national economic or technological security.
• Societal stability: Negligible. The filing concerned a local logging permit dispute and did not threaten social cohesion, civil liberties, or population safety.