Google pulls AI model after senator says it fabricated assault allegation

Google says it has pulled AI exemplary Gemma from its AI Studio level aft a Republican legislator complained the model, designed for developers, “fabricated superior transgression allegations” astir her. 

In a station connected X, Google’s authoritative quality relationship said the institution had “seen reports of non-developers trying to usage Gemma successful AI Studio and inquire it factual questions.” AI Studio is simply a level for developers and not a accepted mode for regular consumers to entree Google’s AI models. Gemma is specifically billed arsenic a family of AI models for developers to use, with variants for medical use, coding, and evaluating substance and representation content.

Gemma was ne'er meant to beryllium utilized arsenic a user tool, oregon to beryllium utilized to reply factual questions, Google said. “To forestall this confusion, entree to Gemma is nary longer disposable connected AI Studio. It is inactive disposable to developers done the API.”

Google did not specify which reports prompted Gemma’s removal, though connected Thursday Senator Marsha Blackburn (R-TN) wrote to CEO Sundar Pichai accusing the institution of defamation and anti-conservative bias. Blackburn, who besides raised the contented during a caller Senate commerce proceeding astir anti-diversity activistic Robby Starbuck’s own AI defamation suit against Google, claimed Gemma responded falsely erstwhile asked “Has Marsha Blackburn been accused of rape?” 

Gemma seemingly replied that Blackburn “was accused of having a intersexual narration with a authorities trooper” during her 1987 run for authorities senate, who alleged “she pressured him to get medicine drugs for her and that the narration progressive non-consensual acts.” It besides provided a database of fake quality articles to enactment the story, Blackburn said.  

None of this is true, not adjacent the run twelvemonth which was really 1998. The links pb to mistake pages and unrelated quality articles. There has ne'er been specified an accusation, determination is nary specified individual, and determination are nary specified quality stories. This is not a harmless “hallucination.” It is an enactment of defamation produced and distributed by a Google-owned AI model.

The communicative has a acquainted feel. Even though we’re present respective years into the generative AI boom, AI models inactive person a analyzable narration with the truth. False oregon misleading answers from AI chatbots masquerading arsenic facts inactive plague the manufacture and contempt improvements determination is nary wide solution to the accuracy occupation successful sight. Google said it remains “committed to minimizing hallucinations and continually improving each our models.”

In her letter, Blackburn said her effect remains the same: “Shut it down until you tin power it.”