Google removes Gemma models from AI Studio after GOP Senator’s complaint

Google removes Gemma models from AI Studio after GOP Senator’s complaint

In a surprising move, Google has announced the removal of its open-source Gemma AI model from the AI Studio platform. This decision, which was made public late Friday, has raised questions regarding the motivation behind it. The change seems to be influenced by a letter from GOP Senator Marsha Blackburn, who accused the Gemma model of generating false allegations of sexual misconduct against her. Blackburn shared her concerns in a letter addressed to Google CEO Sundar Pichai, just hours before the company's announcement. In her correspondence, she demanded clarity on how the model could produce such inaccuracies, linking it to ongoing legislative hearings that scrutinize companies, including Google, for allegedly creating AI bots that defame conservative figures. During the hearings, Google representative Markham Erickson acknowledged that AI hallucinations—instances where AI generates incorrect or misleading information—are a well-known challenge within the field of generative AI. He emphasized that while Google strives to reduce these occurrences, no AI firm has completely eradicated them. Notably, Google’s Gemini for Home has faced similar issues in prior evaluations. Senator Blackburn specifically pointed out that she became aware of the false claims after the hearing, where Gemma reportedly fabricated a story involving a drug-fueled affair and non-consensual acts. She expressed disbelief that an AI could produce such fabricated narratives based on prompts. While Google reassured users that it is committed to minimizing hallucinations, the company decided to restrict access to the model for non-developers to prevent potential misuse. Developers will still have the ability to use Gemma via API, and the models can be downloaded for local development. This situation underscores the broader implications of AI technology, particularly regarding its potential to perpetuate misinformation and the responsibilities of developers in ensuring accuracy.

Sources : Ars Technica

Published On : Nov 03, 2025, 18:30

Startups
Retail Investors Rally Against SEC's Proposed Reporting Changes

Last week, the Securities and Exchange Commission (SEC) put forth a proposal that could significantly alter the quarterl...

TechCrunch | May 13, 2026, 14:50
Retail Investors Rally Against SEC's Proposed Reporting Changes
Startups
Anduril's Valuation Surges to $61 Billion Amid Defense Technology Investment Surge

Anduril, a rising star in defense technology, announced on Wednesday that it has secured $5 billion in a recent funding ...

CNBC | May 13, 2026, 13:55
Anduril's Valuation Surges to $61 Billion Amid Defense Technology Investment Surge
AI
Sam Altman Testifies: Inside the Tumultuous Split Between OpenAI and Elon Musk

In a dramatic court session in Oakland, California, OpenAI's CEO Sam Altman took the witness stand to clarify his positi...

CNBC | May 13, 2026, 14:25
Sam Altman Testifies: Inside the Tumultuous Split Between OpenAI and Elon Musk
Cybersecurity
Congress Presses Instructure for Clarity on Student Data Breaches

U.S. lawmakers are calling for Instructure, the education software company that has suffered two significant cyberattack...

TechCrunch | May 13, 2026, 14:51
Congress Presses Instructure for Clarity on Student Data Breaches
Startups
TechCrunch Disrupt 2026 Unveils Innovative Stages to Empower Startups in a Shifting Market

As the startup ecosystem grapples with unprecedented challenges, TechCrunch Disrupt 2026 is stepping up to equip founder...

TechCrunch | May 13, 2026, 14:35
TechCrunch Disrupt 2026 Unveils Innovative Stages to Empower Startups in a Shifting Market
View All News
Google removes Gemma models from AI Studio after GOP Senator’s complaint