Skip to content
Submit incident
AII-1001-45077
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.Supporting sources are pending verification.

Austrian Supreme Court rejects AI-generated legal complaint with fabricated citations

An Austrian lawyer filed a legal complaint generated primarily by an AI system on behalf of a convicted defendant. The Supreme Court rejected the complaint after discovering it contained numerous false citations and fabricated references to court decisions that did not exist.

3 November 2025 Β· Reliability and hallucination
3 November 2025
AII-1001-45077
Quick facts
Category
Reliability and hallucination
Country
Austria
Date
3 November 2025
Drafted by
Raidu Incident Detection

What happened

The Austrian Supreme Court (OGH) rejected a legal complaint (Nichtigkeitsbeschwerde) filed in a criminal case that was generated primarily using artificial intelligence. The complaint had been filed on behalf of a man convicted by Graz Regional Criminal Court partly for drug trafficking.

Reported impact

The AI-generated complaint was filled with erroneous citations and referenced Supreme Court decisions that did not exist. The complaint cited alleged procedural results and court rulings that the AI apparently fabricated, failing to meet the required legal standard for presenting grounds for annulment at the appropriate argumentation level for Austria's highest court.

Index analysis

The OGH noted that the complaint, created apparently without expert legal review, was saturated with numerous misquotations regarding both alleged procedural results and mostly non-existent or irrelevant Supreme Court decisions. Factually, the complaint referenced case materials at locations where they did not appear in the court file.

Governance controls

The OGH declined to address the complaint on its merits, finding that it did not meet the threshold for legal argumentation expected of Austria's highest court. The court's decision highlights the absence of professional oversight during the AI-generated document's creation.

Classification

Category
Reliability and hallucination
Industry
Government, security and defence
Country
Austria
Harm status
Actual harm
Intent
Unintentional
Timing
Post-deployment
Failure modes
Fabricated citation, Hallucination
Consequences
None recorded
AI systems
Unspecified large language model

Sources and evidence

Sources not yet published

This record was researched and written by the Index from original reporting. Its sources are still being verified by an editor, and they are listed here once that check is complete. Until then the record carries no reviewer.

How sources are verifiedSuggest a source for this incident