~/ARTIFICIAL I/fake-israeli-funded-think-tank-optimized-content-to-influence-ai-models

Fake Israeli-Funded Think Tank Optimized Content to Influence AI Models

A report revealed that a fake US-based think tank called the Hanover Institute for Public Policy, funded by the Israeli government, published articles specifically optimized to be retrieved and cited by AI models. This operation represents a shift towards targeting AI search engines and LLMs for propaganda dissemination. This marks a novel vector in information warfare, transitioning from traditional social media manipulation to Generative Engine Optimization (GEO) to poison the knowledge base of LLMs. As users increasingly rely on AI for information retrieval, such tactics could systematically bias AI-generated answers on geopolitical topics. The operation targeted AI-powered search engines by structuring digital content to rank highly in LLM citations, a technique known as Generative Engine Optimization (GEO). Concurrently, another fake think tank, the International Burke Institute, was reportedly run by Russia to generate pro-Israel credibility using ChatGPT.

## BACKGROUND

Generative Engine Optimization (GEO) is the practice of structuring online content to improve its visibility and citation frequency in AI-generated search results, such as Google AI Overviews or ChatGPT. Additionally, data poisoning attacks on LLMs involve injecting manipulated or biased information into the datasets or web sources that these models retrieve to answer user queries.

## REFERENCES

## KEYWORDS

#Artificial Intelligence#AI Safety#Information Warfare#Large Language Models#Geopolitics

$ subscribe --daily

Fake Israeli-Funded Think Tank Optimized Content to Influence AI Models | Daily News