A federal judge questioned Google’s characterization of its AI Overviews as a product improvement, describing the practice as potentially unfair to publishers during an ongoing antitrust case.
Judge Amit Mehta expressed skepticism regarding Google’s labeling of AI Overviews during a hearing in Penske Media’s antitrust lawsuit against the company, reportedly describing the AI Overviews as “really unfair” to content creators whose work is used.
The judicial scrutiny comes as new reports indicate Google’s recent spam update may have impacted fully automated AI-generated content, further complicating the search giant’s integration of artificial intelligence into its core services.
Jason Kint, CEO of Digital Content Next, previously highlighted concerns about Google’s use of publisher content for AI training, according to earlier reports.
Separately, reports from affected website owners suggested that Google’s August spam update significantly impacted sites relying heavily on fully automated AI content. While Google has not officially confirmed targeting AI-generated material, numerous site owners reported substantial traffic declines.
Google’s Search Relations team, through John Mueller, stated that the company does not require any special actions for generative AI responses in search regarding geographical targeting, according to Roger Montti.
New data from Shero Commerce indicates that while brands are frequently recommended in AI answers, their own official pages receive direct citations only 31 percent of the time. Third-party websites often receive the linking attribution instead, according to Dan Akeju.
Takuma Oka and Aleyda Solís also contributed to the analysis of AI citations and their implications for brand visibility.
Further testing by Promptwatch and Suganthan Mohanadasan revealed inconsistencies in how ChatGPT cites content from Reddit, suggesting that the share of citations does not fully reflect all content considered by the AI model.
Source: Search Engine Journal
