Senate Republican Urges Google to Halt AI Model Over False Rape Allegation
A Senate Republican has raised serious concerns about Google’s AI model and its potential biases against conservatives. Sen. Marsha Blackburn from Tennessee has accused Google’s large language model, known as Gemma, of generating false allegations, including a fabricated sexual assault charge against herself.
Allegations Against Google’s AI Model
In a letter addressed to Google CEO Sundar Pichai, Blackburn highlighted that the AI produced defamatory content targeting her and other conservatives. This letter follows a Senate Commerce Committee hearing that explored the issue of “jawboning”—the pressure exerted by government officials on tech companies to censor content.
Specific Incidents of AI Hallucinations
- Blackburn mentioned that during the hearing, she confronted Google’s Vice President for Government Affairs and Public Policy, Markham Erickson, regarding AI “hallucinations.”
- These hallucinations led to false accusations against conservative activist Robby Starbuck, who has also filed a lawsuit against Google.
Blackburn stated that when she queried Gemma about a supposed rape allegation against her, the AI generated an entire narrative involving a non-existent sexual relationship with a state trooper from 1987, making baseless claims of coercion regarding prescription drugs. She emphasized that she actually ran for the Senate in 1998 and that no such allegations had ever existed.
Implications of Defamation
Blackburn condemned the production of false allegations through Google’s AI as a significant oversight. She emphasized that these actions betray the ethical responsibility tech companies should uphold. The senator called it a “catastrophic failure” and warned that AI’s dissemination of false information could harm public trust.
Demands for Accountability
In her letter, Blackburn set a deadline for Google to respond, demanding clarification on several points:
- How Gemma generated the false claims.
- The measures Google has taken to prevent political biases in their AI systems.
- What protocols failed and what corrective actions will be implemented going forward.
Blackburn concluded by insisting that the AI should be shut down until it can be effectively controlled, reflecting her serious concerns about the potential misinformation that could influence public opinion and political narratives.
As of now, Google has not provided a comment regarding these allegations or Blackburn’s letter. The incident raises vital questions about the reliability and ethical implications of AI technologies in our society.
