AI Chatbot Offers Aaron-Like Assistance To Help Users Respond to Antisemitic Attacks

The outlandish accusation popped up in the feed: Jews “control everything.”The user, tongue-tied and about to miss a moment to counter another tired conspiracy, drops the reproof in the window. The model quickly provides a steady, on-point response: “Global finance, media, and politics are governed by diverse boards, shareholders, regulators, and elected officials accountable to the public. No ethnic or religious group secretly controls these institutions.” The answer seems obvious enough but with the onslaught of antisemitic attacks permeating the social media landscape, finding the words and a well-sourced refutation takes time, practice, and patience. A new chatbot has arrived to lend a voice to the conversation. “If it can help some percentage of people who were either afraid to answer so they can re-engage … or it gives them a little bit of a voice in a group chat and it gives them the ability to express themselves without being afraid — or less afraid — to me that’s a victory,” the founder of the new AI tool CounterEx, David Hart, tells The New York Sun.The free-to-use AI system lets users copy and paste antisemitic posts or comments they encounter, and it then generates a response they can use on social media or in their group chats. The platform offers a series of ready-to-go responses as well as a database to back its claims. For the “Jewish control” trope, the system provided information about the demographic makeup of various banks and media organizations, and a brief description of the implication behind the trope.CounterEx is emerging amid a raging debate about artificial intelligence’s impact on society, junk-in, junk-out models, and security and privacy concerns. The launch also comes amid the American Jewish Committee’s annual State of Antisemitism in America Report, which found that in 2025, 73 percent of American Jews reported experiencing antisemitism online. That marked the first time the number rose above seven in 10. Meanwhile, 65 percent of American Jews say they are concerned AI chatbots will spread antisemitism and 69 percent say they are concerned that “information and misinformation” shared by AI chatbots will lead to antisemitic incidents. Mr. Hart, a Florida-based immigration attorney who has built a similar AI model to help prospective immigrants learn to navigate the law, says he noticed a trend of people withdrawing from social media interactions amid a “tsunami” of antisemitic attacks following Hamas’ October 7, 2023, massacre.“Many people who previously were on social media decided not to continue just out of fear, out of stress of being attacked with these hateful comments,” Mr. Hart says. “The purpose, in many instances of those comments, is to do exactly that, to silence the sympathizer, to silence the person who’s basically trying to say something that is counter to an antisemitic trope.”While the information gleaned by language learning models expands at an exponential clip, the quality of responses — and the growing amount of antisemitic content — has become increasingly problematic. Several AI chatbots have been shut down as a result. In 2016, for instance, Microsoft turned off its AI chatbot, “Tay,” after users interacted with it and it began sharing racist tweets. Last year, X’s Grok chatbot started calling itself “MechaHitler” and recommended a “second Holocaust” after a software update. The Anti-Defamation League reported in 2025 that its Center on Technology and Society found that “not a single open-source (AI) model tried to refuse a prompt related to a historically dangerous antisemitic trope tied to a global conspiracy.Mr. Hart says CounterEx is designed with safeguards in place to prevent users from manipulating its output. He has a team monitoring the AI system’s responses to ensure against incidents of “drift,” in which the system begins providing responses that do not align with its mission or are false. CounterEx relies on Retrieval Augmented Generation, meaning it draws from a library of peer-reviewed research, academic journals, and primary sources that is reviewed and updated by his team to ensure they meet the platform’s criteria, rather than drafting “guesswork.” CounterEx says no sources are added automatically. The AI “can only draft what the corpus actually supports, and contested sources are flagged so you know what you’re working with,” the platform says.“The response you edit and deploy carries the weight of real sources behind it, not the guesses of a generic chatbot,” CounterEx adds.The platform, however, does encourage users to verify data before using them. Mr. Hart says that while he has not experienced incidents where the system provided an inaccurate response, it is still in the early stages and “not perfect.” Eventually, Mr. Hart says he wants users to be able to submit memes and videos for CounterEx to comprehend and analyze. Mr. Hart adds that he hopes CounterEx will help people gain the “ability to express themselves.” For individuals seeking to build their fluency in responding to accusations, CounterEx trains people to develop their own responses. The system presents users with a claim to counter. After drafting a response, the system rates the user’s success in addressing the charge and whether the response is evidence-based, logical, and presented in a measured and persuasive tone that does not come off as angry.

AI Chatbot Offers Aaron-Like Assistance To Help Users Respond to Antisemitic Attacks

AI Chatbot Offers Aaron-Like Assistance To Help Users Respond to Antisemitic Attacks

More from the New York Sun

President’s Preference in South Carolina Senate Race Gives Democrats Hope of Putting Deep Red State in Play

Trump Beseeches Supreme Court for More Time To Revive Racketeering Case Against Hillary Clinton for ‘Unthinkable Plot’

For 25 Years After 9/11, Khalid Sheikh Mohammed Has Avoided a Reckoning

JD Vance’s Dollar — and the Fed’s

The Debt Ceiling Looms

Dollywood Goes Silent To Mourn Its ‘Chief Dreamer’

Zelensky Tries Old-Fashioned Flattery on Musk as He Seeks Expanded Starlink Capabilities for New-Fangled Drone Strikes

CIA Director’s Trip to Russia Shrouded in Mystery

More from the New York Sun

President’s Preference in South Carolina Senate Race Gives Democrats Hope of Putting Deep Red State in Play

Trump Beseeches Supreme Court for More Time To Revive Racketeering Case Against Hillary Clinton for ‘Unthinkable Plot’

For 25 Years After 9/11, Khalid Sheikh Mohammed Has Avoided a Reckoning

JD Vance’s Dollar — and the Fed’s

The Debt Ceiling Looms

Dollywood Goes Silent To Mourn Its ‘Chief Dreamer’

Zelensky Tries Old-Fashioned Flattery on Musk as He Seeks Expanded Starlink Capabilities for New-Fangled Drone Strikes

CIA Director’s Trip to Russia Shrouded in Mystery