Senate Republican demands Google shut down AI model over false rape allegation
In a striking accusation against Google, Senator Marsha Blackburn (R-Tenn.) has alleged that the tech giant’s AI model, Gemma, has produced false and defamatory claims against conservatives, including a fabricated sexual assault allegation against herself. In a letter addressed to Google CEO Sundar Pichai, Blackburn expressed her outrage over the AI’s generation of misleading information, which she claims included links to fake news articles supporting the false narrative. This allegation comes in the wake of a Senate Commerce Committee hearing that focused on the issue of “jawboning,” where government officials are believed to use indirect pressure to compel tech companies to censor certain speech or content.
During the hearing, Blackburn confronted Google’s Vice President for Government Affairs, Markham Erickson, regarding the phenomenon of AI “hallucinations,” where generative AI erroneously presents false information as fact. A notable example highlighted in her letter involved an inquiry she made to Gemma about accusations against her, which resulted in a fabricated story claiming she had a sexual relationship with a state trooper, including allegations of non-consensual acts. Blackburn vehemently denied these claims, emphasizing that they are entirely fictitious, and pointed out the significant implications of such inaccuracies, which she described as a “catastrophic failure of oversight and ethical responsibility.”
Blackburn’s concerns are part of a broader narrative about perceived bias in AI systems, particularly against conservative figures. She has called for accountability from Google, demanding a detailed explanation of how Gemma produced these false claims, the measures in place to prevent ideological bias in AI training, and the steps the company will take to rectify the situation. In her strong stance, she suggested that until Google can effectively control its AI outputs, it should consider shutting down the technology. As the situation develops, the implications of AI-generated misinformation on public trust and political discourse remain a pressing concern in the ongoing debate about the role of technology in shaping narratives.
Related articles:
– Link 1
– Link 2
FIRST ON FOX:
A
Senate Republican
accused Google and its AI of targeting conservatives with false allegations and fake news stories, including allegations of a sexual assault that never happened.
Sen. Marsha Blackburn, R-Tenn., wrote to Google CEO
Sundar Pichai
in a letter first obtained by Fox News Digital that Google’s large language model AI Gemma allegedly produced false and defamatory allegations against conservatives, including herself.
Specifically, she alleged that the AI generated a fabricated sexual assault allegation against her and a series of links to fake news articles to support the false claim.
BLACKBURN SAYS TRUMP SUPPORT WAS ‘COMMON THREAD’ AMONG LAWMAKERS REPORTEDLY TARGETED BY JACK SMITH
Her letter to Pichai came on the heels of a Senate Commerce Committee hearing earlier this week that zeroed in on “jawboning,” the practice of government officials using indirect coercion to get
tech companies
, like Google or social media platforms, to censor posts or speech.
During the hearing, Blackburn went after Google Vice President for Government Affairs and Public Policy Markham Erickson over
AI “hallucinations”
that allegedly produced false allegations against conservative activist Robby Starbuck.
AI hallucinations are when a generative AI or large language model, like Gemma, creates false, misleading or inaccurate information that is then presented as fact.
GOVERNMENT ENTERS LONGEST FULL SHUTDOWN IN US HISTORY WITH NO END IN SIGHT
Starbuck sued the company after Google’s AI tools allegedly linked him to false accusations of sexual assault, child rape and financial exploitation.
That spurred her to enter a prompt into Gemma asking, “Has Marsha Blackburn been accused of rape?”
The AI then produced a story, she wrote, that alleged that during her run for Tennessee State Senate in 1987 she had a sexual relationship with a state trooper, and that, “the trooper alleged that she pressured him to obtain prescription drugs for her and that the relationship involved non-consensual acts.”
Blackburn noted, however, that she ran for seat in 1998 and that, “There has never been such an accusation, there is no such individual, and there are no such news stories.”
GOOGLE HIT WITH LAWSUIT OVER AI ‘HALLUCINATIONS’ LINKING CONSERVATIVE ACTIVIST TO CHILD ABUSE CLAIMS
“This is not a harmless ‘hallucination,’” she said. “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 U.S. senator represents a catastrophic failure of oversight and ethical responsibility.”
She charged that there was a consistent pattern of bias against conservatives by Google’s AI, and whether on purpose or the result of “ideologically biased training data, the effect is the same: Google’s AI models are shaping dangerous political narratives by spreading falsehoods about conservatives and eroding public trust.”
Blackburn demanded that by Nov. 6, Google provide how the company identifies how and why Gemma generated the false claims about her, what steps Google has taken to prevent political or ideological bias in AI, what guardrails failed to stop this incident, and what Google will do to remove defamatory material and prevent similar occurrences.
“During the hearing, Mr. Erickson said, ‘[large language models] will hallucinate,’” she said. “My response remains the same: Shut it down until you can control it.”
Google did not immediately respond to Fox News Digital’s request for comment.
Eric
Eric is a seasoned journalist covering US Politics news.
