In early 2025, the Norwegian Supreme Court identified a legal filing containing fabricated citations and quotes, which had been generated by an AI tool and submitted by an attorney without verification. The court detected the hallucinations during routine review and subsequently updated its lawyer guidance to emphasize professional responsibility and the necessity of verifying AI-generated content. No sanctions were issued in this instance, but the court signaled that future occurrences could lead to disciplinary action.
In early 2025, Norway’s Supreme Court (Høyesterett) received a legal filing containing hallucinated legal citations and quotes, allegedly inserted by an attorney using an AI tool without verification. The fabricated sources were detected during standard review. While no sanctions were issued, the Court updated its guidance to lawyers, warning that AI use does not exempt them from professional responsibility.
Risk classification
- Primary risk domain: 3 Misinformation
- Primary risk subdomain: 3.1 False or misleading information
The AI system generated fabricated legal citations and non-existent legal sources, which constitutes false and misleading information that was subsequently submitted to a court of law.
Additional risk subdomains
- 5.1 Overreliance and unsafe use: The attorney relied on the AI-generated text and submitted the legal brief to the Supreme Court without verifying the sources, demonstrating unsafe use and overreliance.
Causal factors
- Entity: AI
- Intent: Unintentional
- Timing: Post-deployment
The risk was caused by the AI system generating fabricated legal citations (hallucinations) during its post-deployment use, which was an unexpected and unintended outcome of its operation.
EU AI Act risk tier
- Risk tier: 3 Limited Risk
Risk Level 3. Limited Risk: The incident involves the use of a general-purpose AI system (likely a chatbot or text generator) to produce AI-generated content, which falls under the Limited Risk category requiring transparency and verification.
AI system and alleged parties
- AI system: unspecified
- AI purpose: Writing Assistant; Technical Text Generation
- Behaviour type: Assistant
- Alleged developer: Unspecified generative AI developer
- Alleged deployer: Unnamed Norwegian lawyer
- Alleged harmed parties: Unnamed clients represented by unnamed Norwegian lawyer, Supreme Court of Norway, Legal integrity, Høyesterett
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 Negligible
- Child sexual exploitation and abuse: direct Negligible, indirect Negligible
Epistemic
Reported: The report explicitly describes epistemic harm in the form of fabricated legal citations, non-existent laws, and fake quotes generated by an AI tool.
Directly caused: The AI system generated non-existent legal sources and quotes that were incorporated into a formal legal filing.
Indirectly caused: The submission of these fabricated sources forced court researchers to spend time and resources verifying and ultimately debunking them.
People affected
- Occurrences reported: 1
- People reportedly exposed: 2
Potential causes
Management
- Lack of Firm Support: The lawyer did not have a large firm's internal support and guidance.
Technology
- AI Hallucination: The AI tool generated fictitious legal sources and preparatory works.
- Lack of Source Validation: The generative AI model lacked built-in validation for Norwegian legal facts.
Data Inputs
- Plausible but Fake Data: AI training patterns allowed generation of highly realistic but fake citations.
Human Factors
- Failure to Verify Sources: The lawyer did not check the AI-generated sources before submitting the brief.
- Overreliance on AI Tool: The lawyer trusted the AI output without performing standard due diligence.
- Insufficient AI Competence: The lawyer lacked understanding of AI risks like hallucination.
Process and Methods
- Lack of Quality Control: No internal peer review or citation verification process was used.
Regulatory Environment
- Outdated Court Guidelines: The Supreme Court guidelines lacked explicit instructions on AI use.
- Unclear Professional Standards: Lack of explicit ethical guidelines from the Bar Association on AI.
Information quality
- Classification confidence: High
- Reason for confidence: The report from a reputable Norwegian legal news outlet clearly details the incident, including statements from the Chief Justice of the Supreme Court. The facts regarding the AI's hallucination of legal sources and the lawyer's failure to verify them are explicit and uncontested.
- Ambiguities identified: The specific AI model or tool used by the lawyer is not identified in the report.
- Alternative interpretations: None. The event is straightforwardly an instance of AI hallucination in a legal context.
An attorney submitted a civil legal filing to the Norwegian Supreme Court containing AI-generated fake citations. The court detected the hallucinations and updated its lawyer guidelines. The incident represents a minor, localized impact on government operations with negligible national security implications.
- Overall national security impact: Minor
- Scope: Single nation
- Primary target: Norway
- Alleged perpetrator: Individual
Threat characteristics
- Imminence: Long-term. This represents an ongoing professional standards and regulatory challenge rather than an immediate national security crisis.
- Autonomy: Human-controlled. The AI functioned as a writing assistant, and the human attorney retained final control and responsibility for submitting the document.
- Novelty: Established threat. AI hallucinations in legal briefs have occurred in multiple jurisdictions globally, making this an established threat pattern.
Impact by dimension
- Physical security: Negligible. No physical systems, infrastructure, or human safety risks were present in this incident.
- Information security: Negligible. The incident involved an unintentional hallucination by an AI writing assistant in a civil filing, not a systematic or hostile information operation.
- Sovereignty: Minor. The incident involved a filing to the Supreme Court, a core government institution. However, it was an isolated civil case with no foreign interference or systematic disruption to judicial authority.
- Economic security: Negligible. There was no threat to economic stability, strategic industries, or technological competitive advantage.
- Societal stability: Negligible. No threat to social cohesion, civil liberties, or population safety occurred.
.avif){kind=small}