Germany’s OLG Hamm Rules Companies Strictly Liable for AI Chatbot Hallucinations, Rejecting Autonomous AI Defense
On May 12, 2026, the Higher Regional Court of Hamm (Oberlandesgericht Hamm / OLG Hamm) in Germany issued a landmark appellate ruling (case No. 4 UKl 3/25) that establishes a major legal precedent for AI chatbot liability in Europe. The court ruled that companies utilizing AI chatbots for commercial communications are strictly and directly liable for any false or misleading statements generated by the AI, resoundingly rejecting the defense that "AI hallucinations" are autonomous and therefore not attributable to the company.
This decision, combined with the Court of Pistoia's (Italy) March 2026 ruling on AI unfair competition, signals a powerful and converging European judicial trend: deployers bear full responsibility for the outputs of their customer-facing AI chatbots under unfair competition and consumer protection laws.
Case Background: Falsified Medical Qualifications
The defendant in the case was an aesthetic treatment company operating clinics under the name "Dr. Rick & Dr. Nick." The company integrated an AI chatbot on its website to handle customer inquiries and schedule appointments.
During customer interactions, the AI chatbot repeatedly claimed that the clinics' two directors were "specialists in plastic and aesthetic surgery," "specialists in aesthetic medicine," and "specialists in aesthetic treatments." However, under German medical chamber regulations, these specific titles do not exist, and neither doctor held those qualifications. The AI chatbot had fabricated (hallucinated) these professional designations.
The Consumer Association of North Rhine-Westphalia (Verbraucherzentrale NRW) sent a formal warning (Abmahnung) to the clinic. Although the company disabled the chatbot, they refused to sign a cease-and-desist declaration, prompting the consumer group to file a lawsuit under Germany's Act Against Unfair Competition (UWG).
The Court’s Ruling: AI is a Tool, Not a Third Party
The company argued that it should not be held liable because it had only provided correct, verified source data to the chatbot. It contended that the false statements were entirely a product of the AI's autonomous, unpredictable behavior (hallucination).
The OLG Hamm rejected this defense in its entirety. The court ruled that:
- Direct Attribution: An AI chatbot is not an independent third party, but rather a digital tool ("Werkzeug") of the company.
- Strict Operational Risk: Misleading statements made by a chatbot are directly attributable to the operator. It is legally irrelevant whether the AI "hallucinates" or processes correct inputs incorrectly.
- Full Responsibility: Any company that deploys an AI chatbot to interact with the public bears the full risk and responsibility for its outputs.
Because of the fundamental legal importance of assigning liability for autonomous AI hallucinations under unfair competition law, the OLG Hamm explicitly allowed an appeal (Revision) to the Federal Court of Justice (Bundesgerichtshof - BGH).
Enterprise Risk Implications
For legal and risk teams, the OLG Hamm decision confirms that deploying "black box" customer-facing AI tools is a high-risk operational activity. Companies cannot rely on disclaimers or "autonomous AI" defenses to escape liability for misleading advertising, incorrect pricing, or false claims made by their bots. To mitigate these risks, enterprises must:
- Implement robust, continuous monitoring and testing of chatbot outputs.
- Establish strict system constraints and guardrails to prevent chatbots from answering questions outside of a narrow, verified scope.
- Comply with Article 50(1) of the EU AI Act (effective August 2026), which mandates clear disclosure to users that they are interacting with an AI system.
Verbatim Quotes
"Dem folgte das Gericht nicht. Ein KI-Chatbot ist kein eigenständiger Dritter, sondern ein Werkzeug des Unternehmens. Seine Aussagen werden dem Betreiber unmittelbar zugerechnet – unabhängig davon, ob die KI „halluziniert“ oder korrekte Eingangsdaten falsch verarbeitet. Wer einen Chatbot einsetzt, trägt die volle Verantwortung für dessen Ausgaben." (Translation: "The court did not follow this. An AI chatbot is not an independent third party, but rather a tool of the company. Its statements are directly attributed to the operator—regardless of whether the AI "hallucinates" or processes correct input data incorrectly. Anyone who deploys a chatbot bears full responsibility for its outputs.")