No one needs AI to search the Internet, court says in ruling against Google

That could affect both AI search engines and chatbots that are notorious for not properly interpreting original links, a German court has said. he ruled that Google is guilty of misrepresentation in AI Overviews.

A preliminary ruling came at trial by The Decoderwhere two publishers found that Google’s AI Overviews wrongly linked them to fraud and other browsing practices. After mocking the publishers and saying certain words like “Yes, (it) is known for dubious business practices and is often seen as a fraud,” Google failed to correct the errors, even after the publishers sent a cease and desist letter earlier this year.

Google tried the usual arguments to defend itself against the charge of false claims in its AI Overviews, such as saying that most users understand that AI outputs are not always accurate and must be verified.

But the court found that, unlike traditional search engines that only provide third-party word lists, Google’s tool created “independent, innovative, and logical” results based on its misinterpretation of links on the Internet.

This is problematic, the court said, because while publishers can stop others from publishing defamatory terms that appear in Google’s search results, only Google can correct what is listed in AI Overviews. And because, in the first place, the company did not, therefore it “must be sued,” the court ruled. Further, Google’s argument was considered to be too weak, because the AI ​​display in this case “contains words that do not appear in any results.”

The court order—seeking a temporary injunction to stop Google from publishing false claims in its AI Overviews supplement—could have global implications, as the court appears to be the first to sue an AI company over AI speech.

In the past, AI companies hoped that false-warning objections would protect them from lawsuits for being untrustworthy. Last year, a chatbot developer also challenged this AI speech and its “pure speech” class and the First Amendment should protect it.

According to Google’s interpretation of the German court’s decision, however, the false results were “advertisements of the defendant’s marketing,” and the AI ​​tool’s “ideas” and false claims can influence public opinion.