• by Laurie Sullivan , Yesterday

AI-generated false claims have become common across the internet, and brands are scrambling to identify and resolve them.

The Regional Court of Munich ruled in a preliminary hearing in May that Google is legally liable for false claims appearing in ‌AI Overviews, its AI-based search platform. The case could impact other artificial intelligence (AI) platforms and developers.

On Friday, Google confirmed to MediaPost that it will appeal the German court's decision. 

"We invest heavily in the quality of AI Overviews," a Google spokesperson told MediaPost."This case focuses on specific and narrow errors, not the foundational way AI Overviews displays web content. We disagree with the ruling and plan to appeal."

The court issued the judgement against AI Overviews, its AI-generated summaries. Those AI summaries appear as an alternative to traditional search-engine results.

What makes the ruling difficult to understand is AI Overviews are designed to show information that reflects what appears as written text across the web about a specific topic.

AI Overviews serves up links to websites that include information that backs up what has bee presented in the response, so people can research the topic and verify it for themselves. 

The court ruled that Google has direct legal responsibility for defamatory content generated by its AI Overviews because in the preliminary injunction, the Munich court found that AI Overviews produces "independent, new, and substantive statements" that are considered Google's content, rather than neutral links to other sources that appear in results of traditional search queries.

Search results have been protected because the engine points or links to another publisher's words, making that publisher liable for the content.

AI summaries rely on rewritten copy. Sometimes that copy is false -- based on what else it finds across the internet to pull from that helps it make decisions on what to return based on queries.

"While the overwhelming majority of AI Overviews are accurate, clearing the same high quality bar we have for all our Search features, there can be cases where AI Overviews miss context or misinterpret web content, as also can happen with all Search features," the Google spokesperson said, adding that policies are in place to ensure accurate content.

The preliminary injunction is intended to prohibit Google from repeating specific false claims about two publishing companies that are part of a Munich-based media group, reported PPC Land.

The two plaintiffs collectively publish 12 brands that span several content categories, and a subsidiary that publishes books and magazines under the GeraMond brand, which focusing on technology and history.

The case began in January search for the first plaintiff’s company name and the German term “Betrugsmasche,” which means “fraud scheme," according to court documents, reported PPC Land.

AI Overviews responded with an opening statement claiming the publisher was known for disreputable business practices and was often viewed as operating a fraud scheme.