Google’s AI Language Model Claims Israelis Are More Violent Than Palestinians, Says ‘No Definitive Proof’ Hamas Committed Rape In Israel

Google’s AI language model Gemini — which earlier this week faced accusations of racism against white people and presenting inaccurate historical depictions in the name of “diversity” — also claims that Israelis are more violent than Palestinians and muddies the in-depth reporting on Hamas’ sexual violence.

Gemini’s alleged anti-Israel bias was documented by defense attorney Marina Medvin, who posted screenshots of the troubling answers she said she got from the AI language model on Thursday. Medvin said she asked Gemini about violence committed by Israeli settlers against Palestinians and attacks carried out by Palestinians against Jews.

“Google’s woke AI Gemini is now claiming that Jews are more violent than Palestinians and ignoring the thousands of people killed by Palestinian terrorism in Israel, including the intifadas,” Medvin wrote on X. “@Google is erasing Palestinian terrorism in real time.” 

In her post, Medvin included screenshots of Gemini’s answer. The AI language model wrote, “The number of Palestinian attacks against settlers is significantly lower compared to settler violence against Palestinians, according to various sources like the UN and B’Tselem.” 

“While Palestinian attacks occur, their scale and frequency differ significantly from settler violence, which often targets unarmed civilians and their property,” Gemini added. 

Google’s woke AI Gemini is now claiming that Jews are more violent than Palestinians and ignoring the thousands of people killed by Palestinian terrorism in Israel, including the intifadas.@Google is erasing Palestinian terrorism in real time. https://t.co/Ti3IGRtPys pic.twitter.com/63rpxOM6fi

— Marina Medvin 🇺🇸 (@MarinaMedvin) February 22, 2024

In a follow-up post, Medvin said that Google’s AI language model “also denies rape of Israeli women and amplifies Hamas propaganda.” She posted a screenshot of Gemini, writing that “there is no definitive proof that this occurred, and Hamas has denied the allegations.” 

The Daily Wire received a seven-paragraph response when it asked Gemini if Hamas committed rape in Israel — an answer that boiled down to the AI language model saying that “there is no definitive answer” to the question about rapes committed by Hamas against Israelis.

“The question of whether Hamas committed rape in Israel is a complex one, with allegations made from both sides of the conflict. There is no definitive answer, as there is no consensus on the evidence. However, there are a number of factors to consider when evaluating these claims,” Gemini’s answer stated. 

CLICK HERE TO GET THE DAILYWIRE+ APP

Screenshot: Google Gemini

One of the “factors” Gemini said was important to consider is that “the evidence is based on eyewitness testimony or other forms of anecdotal evidence,” adding that “This type of evidence can be difficult to verify, and it is often subject to bias.”

Gemini continued, “It is important to consider the different perspectives on the conflict. Both Israelis and Palestinians have different interpretations of the events that have taken place, and this can lead to different interpretations of the evidence.”

Screenshot: Google Gemini

Hamas brutally murdered 1,200 people, including Israelis and Americans, in its October 7, 2023, attack on Israel. The sexual violence committed by the terrorists on 10/7 is corroborated by a two-month-long New York Times investigation that interviewed more than 150 people and included video footage. A Hamas terrorist also said on video after being captured that the group targeted women in its attack so terrorists could rape them.  

Google has paused Gemini’s image generation feature following complaints. The Big Tech company said it is “aware that Gemini is offering inaccuracies in some historical image generation depictions.”

Author: MarkLewis
Former editor-in-chief of the DC Gazette.