Key facts
- China Central Television has revealed a method to manipulate artificial intelligence.
- The practice is known as "poisoning" in China.
- Generative Engine Optimization (GEO) is used to manipulate AI.
- Fabricated content is flooded onto the internet.
- The goal is to influence AI model citations.
- This manipulation aims to shape investor decisions.
- The practice poses a significant risk to investors.
China Central Television (CCTV) has issued a warning regarding the potential for artificial intelligence to be manipulated, thereby misleading investors. The state broadcaster detailed a practice known as "poisoning" within China, which leverages Generative Engine Optimization (GEO) to influence AI models. This method involves deliberately flooding the internet with fabricated content. The objective is to manipulate the citations that AI models rely on, ultimately shaping investor decisions by presenting false information as factual. CCTV highlighted this as a significant risk to investors and the broader financial markets, indicating a sophisticated method to exploit AI's reliance on vast datasets for information synthesis and decision-making support.