German court holds Google accountable for defamatory AI hallucinations

Image: Adarsh Chauhan @dyno8426, via Unsplash

A regional court in Bavaria has issued a landmark ruling that could make tech companies legally liable for the content produced by their AI systems.

The case was brought by two publishers in Munich, Verlagshaus24 and a subsidiary, after Google’s AI Overviews feature falsely linked the companies with unsavoury business practices, scams and subscription traps. The court said the tech giant’s AI summaries had mixed up information from genuinely shady companies and invented links that didn’t exist.

In the preliminary injunction handed down on 28 May by the Munich Regional Court, Google is barred from publishing false information about the publishers. It faces a fine of up to €250 000 for each instance of non-compliance, or up to six months’ imprisonment for Google managers. 

According to the ruling, unlike a regular Google search, AI generated summaries should be considered the tech company’s own content. “Since the defendant introduced the artificial intelligence itself and offers it to users, it must also be held accountable for its results, as it alone has influence over the AI’s offering and the algorithms it uses,” the court found.

The ruling differentiated AI generated summaries from classic search engine results, which are protected from direct liability by existing federal case law in Germany. But the court said this did not shield Google from being held liable for false information generated by its AI systems. “The disputed display of the ‘results with AI’ is not a mere display of search results, but rather content attributable to it,” the court found.

The court also rejected Google’s reliance on its right to freedom of expression. It argued that “the expressed opinion was primarily generated by AI and is therefore not an expression of the expressing individuals’ own convictions, but rather the result of an algorithm”. Its AI summaries should be regarded as “primarily an expression of their business activities”.

The decision could have far-reaching implications for government regulation of AI. Although current EU legislation does require tech companies to assess and mitigate systemic risk, including the spread of misinformation and inherent algorithmic biases, it does not provide for direct accountability. This ruling has led to renewed calls for a legal accountability framework for contested information produced by generative AI tools.

The Munich ruling could also become a major headache for tech companies who are currently not held legally liable for AI hallucinations. If upheld, it could spark a flood of lawsuits worldwide from companies and individuals who believe they have been defamed by AI generated statements and decide to sue for damages.

Google says it will oppose the judgement. “This case focuses on specific and narrow ​errors, not the foundational way AI Overviews displays web content,” the company told Reuters in an email. “We disagree with the ruling and plan to appeal.”