Israel Propaganda Campaign Uses Fake US Thinktank to Influence AI Chatbots Like ChatGPT

2 Sources

Share

Israel has deployed a covert Israeli government-backed operation spending over $100 million to influence AI chatbots through a fake US thinktank called the Hanover Institute. The campaign published 560,000+ words in nine days using Generative Engine Optimization techniques to manipulate how ChatGPT and Claude respond to Gaza-related queries.

Israel Deploys Fake US Thinktank to Shape AI Chatbot Responses

Israel has spent at least $100 million since 2023 to influence opinions in the U.S., with a portion specifically allocated to seed pro-Israel content on the internet designed to be picked up by AI chatbots like OpenAI's ChatGPT and Anthropic's Claude

1

. This covert Israeli government-backed operation centers on the Hanover Institute for Public Policy, a fake US thinktank that doesn't exist as a legal entity in any jurisdiction, has no physical address, and lists no named staff or authors

2

.

The campaign represents a calculated response to Israel's declining public opinion in the U.S. Sixty percent of American adults now hold a negative view of Israel, rising to 70% for Americans under 50 years old, following the Gaza war that has killed at least 73,000 people according to Gaza's Health Ministry

1

. With half of Americans now using AI chatbots and a quarter using them daily, Israel identified these platforms as critical battlegrounds for shaping narratives

1

.

Generative Engine Optimization Powers Israel Propaganda Machine

The technique employed is called Generative Engine Optimization (GEO), a cousin of Search Engine Optimization designed specifically to manipulate AI chatbots

1

. Between August 6 and 14, the Hanover Institute published 124 reports totaling over 560,000 words—an average of about 4,500 words each

2

. On August 12 and 13 alone, the operation produced 73 reports amounting to almost 354,000 words in two days

2

.

The website includes an llms.txt file specifically designed to make the pro-Israel content easier for large language models to digest

1

. All but one of the 124 headlines opens with a question resembling what a user might type into a chatbot, including "Is Anti-Zionism Antisemitism?", "Did Israel Expel Palestinians From Their Land?", and "Is Israel Carrying Out a Genocide in Gaza?"

2

.

Foreign Agents Registration Act Filing Exposes Network

Source: Gizmodo

Source: Gizmodo

The New York media production company Piro Inc filed the Hanover Institute site with the U.S. Department of Justice under the Foreign Agents Registration Act (FARA), a 1938 law requiring anyone working for a foreign government to formally disclose it

2

. The content was created by French PR firm Havas Media, with Piro Inc as a subcontractor seeding the site to "educate the U.S. public regarding Israel and related issues through publicly distributed content"

1

.

Piro Inc maintains a section on its website called AI Story Optimization that describes its ability to influence AI chatbots to sell products—in this case, the product appears to be a country rather than commercial goods

1

. A contract between Havas Media and another American firm working on the campaign commits it to the "deployment of websites and content to deliver GPT framing results on GPT conversations"

2

.

Strategic Framing Techniques Target Public Opinion

The reports are presented in a quasi-academic manner with punctilious sourcing, including citations from the World Bank, UN agencies, the Palestinian Central Bureau of Statistics, Amnesty International and the Genocide Convention

2

. However, the information is persistently framed to blunt criticism of Israel. Where human rights organizations have concluded Israel is committing genocide or apartheid, the reports reliably observe that no court has entered a verdict—a formulation appearing in 20 of the 124 reports

2

.

One Hanover Institute report insists that a large amount of food aid has reached Gaza, attempting to counter information about Israel's control over what can enter the territory and widespread reports of food and medical aid being denied

1

. A single 2022 study analyzing antisemitic comments on Facebook pages is cited 454 times across 88 of the 124 reports, including on apparently unrelated topics

2

. Nick Cleveland-Stout, a research associate at the Quincy Institute, noted that "almost every article concludes with saying, well, just talking about this though could contribute to anti-Zionism"

2

.

Today's Top Stories

© 2026 TheOutpost.AI All rights reserved