Israel-backed think tank accused of flooding the web to win citations from ChatGPT and Perplexity

An Israel-funded think tank, the Hanover Institute for Public Policy, has been accused of publishing large volumes of unsigned content with one clear goal: getting cited by AI systems such as ChatGPT and Perplexity. According to the report, Hanover has posted more than 100 articles on Israel-Palestine issues since Aug. 6, many framed as direct questions users might ask an AI assistant. The site also provides an llms.txt file that organizes its pages in a machine-readable index. Politico said that when it tested neutral prompts, both ChatGPT and Perplexity repeatedly cited Hanover. Separate checks by 404 Media and Responsible Statecraft found that most sampled articles were flagged by AI detectors as AI-generated. The report also points to Piro, a public relations firm that openly markets an "AI Story Optimization" service designed to identify the pages most likely to shape model responses and then produce material more likely to be cited by large language models. Piro is now using Politico’s test results to promote that service. U.S. FARA filings show Piro worked through Havas Media on a communications project for Israel’s Government Advertising Agency, LaPam, with two disclosed payments totaling $1 million.