German Court Holds Google Accountable for Fake AI-Generated Information
A German court ruled that Google's AI Overviews are the company's own statements, not merely a retelling of search results. This decision could have significant implications for all artificial intelligence developers.
The case was heard by the Munich Regional Court. It stemmed from an incident where AI Overviews erroneously linked two Munich publishers to fraud, "subscription traps," and questionable business practices. However, the sources referenced by the search did not contain such claims — the artificial intelligence independently combined information about other companies and attributed it to the plaintiffs.
The court noted that a traditional search engine merely displays links to third-party content, while AI Overviews creates new text, analyzing and interpreting information "in its own words." Therefore, Google itself bears responsibility for such statements.
The company's argument that users can verify the information in the links themselves was rejected by the court. According to the judges, the fact that a person can conduct additional research does not absolve the party disseminating a false statement of responsibility.
The court also disagreed that the same rules of limited liability applicable to conventional search engines should apply to AI Overviews. In the court's view, AI does not simply find third-party information but creates its own informational product.
Separately, the court addressed the issue of freedom of expression. The ruling noted that an opinion generated by an algorithm is not a personal conviction of a human, and therefore has a lower degree of protection compared to human expressions.
As a result, the court prohibited Google from disseminating a series of false statements about the publishers and ordered the company to cover 80% of the legal costs. The court also noted that even the removal of specific answers does not guarantee that the algorithms will not create similar false statements again.
Google stated that it invests significant resources in improving the accuracy of AI Overviews and is reviewing the court's decision, which is not yet final. The company emphasizes that AI can sometimes lose context or misinterpret website content.
This decision could become an important precedent not only for Google. If such legal logic gains support in higher courts or other countries, similar claims could also affect other generative AI providers that create responses based on internet materials.
Comments
Желание модерировать и наживаться на ресурсах, созданных в свободных странах - бесконечно.