A regional court in Munich and a substantial number of Alphabet Inc. investors agree: AI-generated misinformation on Google’s platforms presents real risks to the public and to the company.
The investor voice was heard on June 5, at Alphabet’s annual shareholder meeting, through a vote on a shareholder proposal filed by Vancity Investment Management, a Vancouver-based firm, with support from Open MIC, which cited research showing that Google’s search AI Overview — now used daily by billions of people — spreads false claims nearly 17% of the time. The proposal cited multiple examples of Google’s AI spreading false information, increasingly leading to lawsuits and regulatory concern.
That shareholder proposal won 28.9% of the independent shareholder vote — excluding the founders' supervoting shares — representing $422.2 billion in market capitalization. A shareholder vote over 20% generally demonstrates substantial concern about a company’s policies and practices, with many institutional investors and boards viewing it as a mandate for dialogue and engagement.
In addition to material risks posed to Alphabet, the shareholder proposal argued that the proliferation of AI related falsehoods by platforms like Google is creating a larger problem for society, what has been called “epistemic collapse” – a world in which users are increasingly unable to discern what is true or authentic.
The truth and authenticity of Google’s AI Overviews was likewise at the heart of a landmark Munich court ruling in early June, which hit the company with a temporary injunction barring it from spreading false claims about two Munich-based publishers through its AI-generated search overviews. The court reportedly found that AI Overviews had falsely tied the publishing companies to “scams, subscription traps, and shady business practices for certain search queries.” Google’s AI mixed up information about other, genuinely sketchy companies with the plaintiffs and drew connections that didn't appear in any of the linked sources, according to the court.
The court ruled Google’s AI Overviews represent the platform’s own answers, not search results. According to published reports, the court concluded that AI Overviews generated independent, new, and substantive statements that were “in its own words and according to its own structure” rather than, like a conventional search engine, simply pointing users to outside websites.
Google, in a statement to the Decoder, an industry publication, says its AI overviews are designed to "reflect" information that already exists on the web and that “the overwhelming majority” of AI Overview responses provide accurate information. Google said it is reviewing the court decision, which is not yet final.