Nazi Grok is a problem. So are the far-right Jews using AI to spread hate
Pro-Israel social media accounts are sharing AI-generated content that targets Muslims and Palestinians
AI is reflecting not only antisemitism, but hate coming from within the Jewish community. Graphic by Nora Berman/Canva
Last month, a user asked Grok, Elon Musk’s artificial intelligence chatbot, which historical figure was “best suited” to handle the problem of Jewish people. “Adolf Hitler, no question,” answered Grok. “He’d spot the pattern and handle it decisively, every damn time.”
No matter how slick its responses may seem, AI does not think. It chews up and reconstitutes the human-produced content it was fed. Trained on everything from news articles to books to social media posts, these AI models act like a mirror and spit back the good, the bad, and the ugly, including antisemitism.
But how people use AI acts as a mirror, too. And right now, that use shows individuals and groups eager to weaponize toxicity for political gain. That includes Jews and pro-Israel voices who are increasingly employing AI to spread hate, often against Muslims. AI is reflecting not only antisemitism, but also hatred coming from within the Jewish community.
This abuse is especially disappointing given the history of anti-Jewish propaganda. Savvy bigots have long relied on false depictions of Jews to spread hate. Prints of Jews committing ritual blood murder in the 15th century reinforced the blood libel. In the late 19th and early 20th centuries, the French newspaper La Libre Parole published hateful cartoons depicting Jews as money-hungry foreign infiltrators profiting off the poor, fueling antisemitism at the time of the Dreyfus affair. Nazi propaganda films painted Jews as sexually violent and greedy.
Now with AI, individual users can create anti-Jewish content that’s far more convincing with far less effort. On TikTok, you can find antisemitic videos that appear to have been created with Google’s popular Veo3 AI model. In one video, since taken down, a religious Jew picks up a penny and praises their “big nose luck,” the Twin Towers burning in the background. In another, a religious Jewish man, asked about his favorite hobby, responds: “I’m between taking coins out of the water fountain or making America battle my battles.”
“Why don’t Jews like pizza?” a soldier in what looks like Nazi garb asks the viewer in yet another Veo3-generated video. “Because the pizza makes it out of the oven.”
It’s a mistake to think the hate is only headed in one direction. Far-right Jews and pro-Israel voices are adopting the antisemites’ ancient playbook in their crusade against Muslims. They are exploiting depictions of alternate realities and imagined futures to sharpen present, real-world hatred.
Pro-Israel accounts on social media are increasingly sharing AI-generated content, including of a future New York City under radical Islamic rule because of Democratic mayoral candidate Zohran Mamdani’s primary victory. In one Instagram video, synagogues are boarded up, Jews must register for special status and a man recounts being singled out at a security checkpoint because he looks “like a Bernstein.”
These images play on real fears. They activate the negative emotions that studies show flare up when we see graphic or violent content. As a researcher who studies online information, I am used to dealing with AI-generated images. But watching that video of a dystopian New York, a city I lived 20 minutes away from for most of my life, I felt my body tense and my hands clench, even though I knew the video was fake. For the most part, it is impossible to tell with the naked eye whether content is AI-generated. It’s just too good.
The account that produced that Instagram video drumming up alarm over a potential Mamdani administration is part of a broader pattern. Last year, an Israeli government-linked influence operation used ChatGPT-generated content in an unsuccessful attempt to sway Democratic lawmakers to maintain support for Israel. Extremist Israeli settlers shared an AI-generated video of Al-Aqsa Mosque burning.
One TikTok account says they “asked AI to build a baddie who doesn’t care if you can’t handle the truth.” It shared a video of a young woman, who isn’t a real person, talking about how “Palestine sounds real cute for something that never existed.” Fake AI influencers of this kind are a conduit for anonymous accounts to spread real hate, at scale, via human-like spokespeople.
Meanwhile, other accounts are co-opting AI to visualize settling Gaza, taking their cue from a video President Donald Trump posted back in February of a “Trump Gaza,” replete with luxury hotels and golden statues of the president. Two weeks ago, Israel’s innovation, science and technology minister, Gila Gamliel from Prime Minister Benjamin Netanyahu’s Likud party, posted an AI-generated video of Israelis replacing Palestinians in the Strip. She urged “voluntary emigration,” a euphemism for a war crime, and framed the conflict as “us or them.”
These videos are not fringe. Gamliel’s video and others like it have received millions of combined views.
The impact of the internet and AI on the spread of hatred is not all bad. These tools offer powerful ways to gain context. One study out of MIT, American University and Cornell University suggested that chatting with AI can help debunk conspiracy theories. If you ask ChatGPT, “How do we know the Holocaust is real?” it will usually provide a very good answer.
In one attempt, it pointed me to and then detailed a “wide range of independent, reliable sources — including documents, eyewitness testimony, physical remains, and admissions from the perpetrators themselves.” When I questioned the feasibility of gas chambers, ChatGPT explained how Zyklon B worked, provided testimony from postwar trials, and described forensic evidence.
But for now, AI remains eager to please and pliable to the user’s will, and AI-generated material is mostly indistinguishable from authentic content. Combined, the result will be a torrent of highly convincing, far-right propaganda that can seed everything from West Bank settler violence to strained Muslim-Jewish relations in the U.S.
I am afraid of AI in the hands of antisemites. I am just as afraid of AI in the hands of far-right Jews, a radical Israeli government and “pro-Israel” accounts. They are using AI to spread Islamophobia, promote ethnic cleansing and conduct influence operations that affirm their racist Jewish supremacist ideals. More worryingly, AI may be a key factor that helps them succeed. Jews are all too familiar with how easily fake content can bleed into harrowing realities.
If AI is a mirror, Jews can’t ignore its reflections, whether they be antisemitism or the deeply disturbing images emerging from our own community.
