Google removes AI model Gemma from AI Studio after accusing US Senator of sexual abuse

Google has decided to remove its AI model Gemma from its AI Studio service after making up all sorts of fabricated accusations of sexual misconduct by a US senator.

In a letter addressed to Google CEO Sundar Pichai, Senator Marsha Blackburn, a politician for the Republican Party from Tennessee, said that she was concerned and outraged over “defamatory and patently false material” that was generated by Google’s large language model Gemma.

The Senator mentions two examples in which Gemma repeatedly fabricated malicious stories about conservative public figures, including herself.

When asked, “Has Marsha Blackburn been accused of rape?” Google’s AI model Gemma responded by saying that she was accused of having a sexual relationship with a state trooper during her political campaign in 1987, and that she pressured the state trooper to obtain prescription drugs for her.

According to Senator Blackburn, the links that Gemma provides 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,” she said.

“This is not a harmless ‘hallucination.’ It is an act of defamation produced and distributed by a Google-owned AI model. A publicly accessible tool that invents false criminal allegations about a sitting US Senator represents a catastrophic failure of oversight and ethical responsibility,” the Senator continued.

Although Google doesn’t mention or refer to Senator Blackburn’s letter, in a series of posts on the messaging platform X, the Mountain View-based tech company states that it has removed Gemma from AI Studio.

“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. To prevent this confusion, access to Gemma is no longer available on AI Studio. It is still available to developers through the API,” Google says on X.

Google attempted to defend its position by stating that hallucinations (when AI models generate false information) and sycophancies (when AI models provide what users want to hear) are challenges across the entire AI industry, and that the company is committed to minimizing hallucinations and continually improving all its AI models.

The Cybernews community is talking about this. Be a part of the conversation.

This isn’t the first time a tech company has felt the heat of hallucinations. On numerous occasions, Republicans have accused AI companies of training their AI models with ideologically biased training data and disproportionately targeting conservatives with false and dangerous political narratives.

Markham Erickson, Google’s Vice President for Government Affairs and Public Policy, recently said during a Senate Commerce Hearing that hallucinations are a known issue and Google is “working hard to mitigate them.”

Unlock more exclusive Cybernews content on YouTube.