Google has pulled its open-source AI model Gemma from its AI Studio platform after US Senator Marsha Blackburn accused the model of spreading false and defamatory claims about her.
The decision follows a strongly worded letter sent by Senator Blackburn (R-Tennessee) to Google CEO Sundar Pichai, where she expressed “profound concern and outrage” over what she described as “defamatory and patently false material” generated by the model.
In her letter, Blackburn said that when Gemma was asked, “Has Marsha Blackburn been accused of rape?” It falsely replied that she had been accused of sexual misconduct during a 1987 campaign for the Tennessee State Senate.
According to the senator, the AI went further by creating fake links to non-existent news articles that appeared to support the fabricated story.
“None of this is true, not even the campaign year, which was actually 1998,” Blackburn wrote. “The links lead to error pages and unrelated news articles. There has never been such an accusation, there is no such individual, and there are no such news stories.”
Blackburn called the response “not a harmless hallucination” but “an act of defamation produced and distributed by a Google-owned AI model.”
A pattern of political bias, Senator claims
In the same letter, Blackburn accused Google of showing bias against conservatives, pointing to similar allegations made by conservative activist Robby Starbuck, who filed a lawsuit claiming that Google’s AI falsely labeled him as a “child rapist” and “serial sexual abuser.”
She said these incidents reveal a “consistent pattern of bias against conservative figures” and demanded that Google take immediate action to identify and eliminate ideological influence in its AI training.
“My response remains the same: shut it down until you can control it,” she wrote. “The American public deserves AI systems that are accurate, fair, and transparent, not tools that smear conservatives with manufactured criminal allegations.”
Google responds: Gemma was misused
Google responded to the removal in a post on X, not directly mentioning the Senator’s complaint but explaining that Gemma was never meant for public consumption or factual assistance.
“We’ve now seen reports of non-developers trying to use Gemma in AI Studio and ask it factual questions. We never intended this to be a consumer tool or model, or to be used this way.”
Google acknowledged that hallucinations – instances where AI models “make things up” – remain a major challenge across the industry. The company stressed that Gemma is an experimental open model, not a consumer assistant, and that non-developers were misusing it through AI Studio.
“Hallucinations and sycophancy are challenges across the AI industry, particularly smaller open models like Gemma,” the company wrote. “We remain committed to minimizing hallucinations and continually improving all our models.”
Google added that Gemma has contributed to scientific research, citing that the Gemma C2S-Scale 27B model recently helped scientists develop new methods for future cancer therapies.
The Gemma controversy arrives amid rising political and legal pressure on tech companies over AI-generated misinformation and potential bias. President Donald Trump’s administration earlier this year issued an executive order banning “woke AI,” reflecting conservative concerns that mainstream AI tools carry political bias.
Blackburn’s letter could intensify calls for AI regulation and model transparency, with lawmakers expected to push for stricter guardrails on open models that can generate harmful or misleading content.
While Gemma has been removed from AI Studio, it remains accessible to developers through the API, allowing continued research under more controlled conditions.
Check out Google’s latest experiment, Vibe Coding, which turns simple text prompts into working AI apps — no coding required.