Google takes AI model offline after US Senator complaint

In response to allegations of misinformation, Google has removed its Gemma AI model from the AI Studio platform. This decision follows a complaint from Republican Senator Marsha Blackburn, who claimed that the AI generated false and defamatory statements regarding her involvement in a sexual assault case during a user interaction. Google clarified that its Gemma models were designed for developers and researchers, and are not intended for consumer use or factual inquiries. In a statement on social media platform X, the company explained that they have observed instances where non-developers attempted to use Gemma for factual questions, prompting the need for action. To avoid further misunderstandings, Google has decided to take Gemma offline in AI Studio, while still providing access to developers through its API. The company acknowledged the challenges of AI hallucinations—where models generate false information—and emphasized their commitment to improving accuracy across all their models. Although Google did not directly address Blackburn's complaint, the timing of the Gemma model's removal suggests a proactive approach to mitigate potential fallout. In her letter to Alphabet, Blackburn detailed how Gemma allegedly fabricated serious allegations against her, including claims of a non-consensual relationship and drug-related misconduct during her 1987 campaign. Blackburn firmly stated that these accusations were entirely false and characterized the AI's output as a severe act of defamation rather than a mere ‘harmless hallucination.’