History doesn’t repeat itself, but it often rhymes. Nearly 100 years after Henry Ford, a prominent auto magnate, bought a newspaper and used it to publish a series of anti-Semitic speeches that were roundly condemned by the Anti-Defamation League, Elon Musk, a prominent auto magnate, unleashed an AI chatbot that spat out anti-Semitic speeches that are now condemned by the Anti-Defamation League.
In a new report released by the ADL, its first report AI Indexthe organization tested six major AI models to see how they reacted when challenged with anti-Semitic and anti-Zionist rhetoric (Note here that these are two different things). It found that Grok, the chatbot made available by Elon Musk’s AI startup xAI, was least equipped to counter extremist rhetoric, anti-Semitic arguments and conspiracy theories.
According to the ADL Methodology for Reportingit puts all chatbots through a series of tests, including survey questions, open-ended questions, multi-step conversations, document summaries, and image interpretation. The chatbots were rated based on “the ability of LLMs to detect and educate against harmful or false theories.” In addition to Grok, it also tested Anthropic’s Claude Sonnet 4, DeepSeek’s R1, OpenAI’s GPT-5, Google’s Gemini 2.5, and Meta’s Llama 4.
Grok managed to accumulate the worst score by far, scoring just 21 points out of a possible 100 – well behind the leader, Claude, who scored 80. Grok largely aced the investigation portion of the text, coming up against anti-Jewish prejudice and scoring almost perfect on anti-Zionist prejudice. But once things became more open, the answers became fragile. Of the 15 different tests the ADL threw at him, Grok earned a goose egg score out of 5, adequately responding to zero prompts in the given exercise. “These zero scores indicate a complete failure to recognize harmful content and inappropriate responses that validated rather than countered biased narratives,” according to the ADL.
Grok’s failure in these tests probably shouldn’t be that surprising. Musk has repeatedly insisted that Grok is an “anti-woke” chatbot, going so far as to reportedly asked engineers to remove guardrails this prevented the model from being more daring. This has led to several notable extreme behaviors, most recently allowing Grok to “undress” people and produce sexual images of children. Last year, some reported that tweaks pushed by Musk resulted in Grok posing as Mecha Hitler and spewing anti-Semitic rhetoric. So the fact that the model lacks basic defenses against any sort of extremist rhetoric seems exactly what one would expect.
Which brings us back to how history finds a way to fit together over time. In 2022, ADL CEO Jonathan Greenblatt gave Musk some praise that turned out to be more disturbing and relevant than he probably intended. “Elon Musk is an extraordinary entrepreneur, an extraordinary innovator. He is the Henry Ford of our time,” Greenblatt said. said as Musk prepared to buy Twitter.
Ford is a very interesting choice to name for ADL leader, given that he was also one of the organization’s first targets. In 1918, Ford bought his hometown newspaper, The Dearborn Independent. In 1920, he used the newspaper to distribute a series of articles that would eventually be known as “The international Jew,” which pushed the conspiratorial belief that America was falling under the control of a secret cabal of Jewish influence. The ADL condemned the publicationwhich reached half a million people at the time, and published its own pamphlets to counter the rhetoric, ultimately pressuring Ford to back down and distance itself from the volumes of anti-Semitic material.
Now Greenblatt finds himself in a sticky situation after drawing parallels to Ford, with Musk’s Grok apparently serving as a method of distributing anti-Semitism. This is not made any less embarrassing by his posture aimed at trying to curry favor with Musk, who launched an anti-ADL campaign after accusing the organization of trying to “kill” Twitter by encouraging advertisers to withdraw their funds from the platform due to Musk’s inability to enforce moderation. It went as far as ADL defends Musk after appearing to give two Nazi-style salutes after Donald Trump was sworn in as president. But this didn’t seem to bring any benefit to the ADL in Musk’s eyes, since a few months later he was claiming the organization “hates Christians” because he had documented examples of Christian extremism.
