Ddos 
Gemma is an open-source artificial intelligence model developed by Google, available to all developers through Google AI Studio and other distribution channels. However, developers may now find it difficult to locate Gemma within the platform, as Google has quietly removed the model.
The company has not provided an explanation for this decision, but its removal came shortly after U.S. Republican Senator Marsha Blackburn sent a letter to Google CEO Sundar Pichai, demanding clarification as to why the model had generated false and defamatory claims about her.
In her letter, Blackburn also questioned why such failures occurred and linked the incident to ongoing congressional hearings accusing Google and other tech companies of developing chatbots biased against conservatives.
During the hearing, Malcolm Erickson, a representative from Google, explained that AI hallucinations are a common and well-documented issue within generative models. He emphasized that Google is working diligently to reduce the impact of such errors, though no company has yet succeeded in eliminating them entirely.
In her public statement, Blackburn claimed she only became aware of the defamatory response after the hearing. When asked whether she had been accused of rape, the Gemma model allegedly responded by fabricating a story about an extramarital affair with a state trooper under the influence of drugs, even generating fake news links to support the claim—sources that, in reality, did not exist and were entirely hallucinated by the model.
Such incidents are not uncommon in the realm of generative AI. With deliberate prompting, these systems can easily be led into producing false or misleading information, especially when used within environments like Google AI Studio, which allows users to manipulate model behavior. This flexibility, while valuable for experimentation, also makes it easier for bad actors to weaponize AI models for spreading misinformation.
Related Posts:
- U.S. Senators Request Facebook and Google to Attend a Hearing to Answer Platform Security Issues
- Pro-Russian Hacktivist Group TwoNet Exposed for Fabricating Critical Infrastructure Attacks to Boost Reputation
- US Threatens UK Over iCloud Backdoor Demand
- Meet Gemma 3 270M: Google’s Most Lightweight AI Model Yet
- Google Unleashes Gemma 3n: Breakthrough On-Device Multimodal AI for Smartphones & Laptops
Donate