Israel-Backed Fake Think Tank Published 560,000 Words in Nine Days to Influence ChatGPT About Gaza

A pro-Israel website posing as a neutral research institute, the Hanover Institute for Public Policy, published more than 560,000 words in nine days as part of an Israeli-funded effort to shape AI chatbot responses about Gaza, according to a Guardian analysis. The site, registered with the US Justice Department as work for the Israeli government, produced 124 reports between 6 and 14 August, with headlines phrased as chatbot queries, and was designed to be cited by AI systems like ChatGPT, Perplexity, Claude, and Gemini. Researchers say the reports consistently frame information to blunt criticism of Israel, and the site has stopped publishing since scrutiny began.

Israel-Backed Fake Think Tank Published 560,000 Words in Nine Days to Influence ChatGPT About Gaza Hanover Institute's AI-targeted content raises ethical concerns A pro-Israel website posing as a neutral research institute published more than 560,000 words in nine days as part of a wider Israeli-funded effort to shape what AI chatbots say about Gaza. The site, called the Hanover Institute for Public Policy, presents itself like an academic think tank but appears to have no legal existence, physical address, named staff or report bylines. Its reports cover highly sensitive subjects, including whether Israel is committing genocide in Gaza, the torture of Palestinian prisoners, Israeli war crimes and the starvation of Palestinians. The material is written in the style of neutral policy research, with long citations to sources such as the World Bank, UN agencies, Amnesty International and the Genocide Convention. But the site was not an ordinary research project. New York media company Piro Inc registered material distributed through the site with the US Justice Department under the Foreign Agents Registration Act as work for the Israeli government. 124 Reports in Nine Days A Guardian analysis found that between 6 and 14 August, the Hanover Institute published 124 reports totalling more than 560,000 words. On 12 and 13 August alone, it produced 73 reports, amounting to nearly 354,000 words in just two days. Almost every headline was phrased like a question someone might type into a chatbot. Examples included 'Is Israel Carrying Out a Genocide in Gaza?', 'Is Anti-Zionism Antisemitism?' and 'Did Israel Expel Palestinians From Their Land?' The site has reportedly stopped publishing since scrutiny of the operation began. Built for Chatbots, Not Just Readers The most striking part of the project is how clearly it appears designed for artificial intelligence systems. The Hanover site initially carried an llms.txt file, which tells AI systems how to read and interpret a website. That file did not describe the institute, but instead belonged to Res, an 'AI-native content platform' that advertises helping clients get cited by ChatGPT, Perplexity, Claude and Gemini. Res founder Hai Tran registered with the US Justice Department in July as a subcontractor to Piro on the Israeli account. Piro itself advertises 'AI Story Optimization', promising content 'engineered for how LLMs evaluate credibility'. Framing Gaza for Machines Researchers say the Hanover reports consistently frame information in ways that blunt criticism of Israel. Where human rights organisations have accused Israel of genocide or apartheid, the reports repeatedly note that no court has entered a final verdict. Casualty and aid figures are often described as coming from 'a party to the events', a caveat applied broadly enough to leave readers with the impression that little can be firmly established. Nick Cleveland-Stout of the Quincy Institute told The Guardian the site initially looked credible. 'It succeeds in masquerading as very academic,' he said. 'But the more you read it, you're able to realise that a lot of it makes no sense.' Wider Israeli-Funded Campaign The Hanover Institute is part of a broader campaign funded with tens of millions of dollars from the Israeli government and routed through third parties, including Havas Media. US filings show Havas-linked contractors included Piro and Clock Tower X, a firm run by former Trump campaign manager Brad Parscale https://www.ibtimes.co.uk/israel-ai-influence-campaign-chatbots-1812371 . One contract reportedly described the goal as deploying websites and content to deliver 'GPT framing results' in chatbot conversations. The Israeli foreign ministry has said it does not conduct influence operations in the US and follows American law. Piro co-founder Daniel Rosenberg has said the company was hired to put accurate, sourced facts into the public record and counter misinformation about Israel. © Copyright IBTimes 2026. All rights reserved.