How to Spread Disinformation with Unicode

There are many different ways to represent the same text in Unicode. We’ve previously exploited this encoding-visualization gap to craft imperceptible adversarial examples against text-based machine learning systems and invisible vulnerabilities in source code. In our latest paper, we demonstrate another attack that exploits the same technique to target Google Search, Bing’s GPT-4-powered chatbot, and … Continue reading How to Spread Disinformation with Unicode →

academic papers adversarial attack bing chatbot code disinformation encoding exploited exploits gap google google search gpt gpt-4 latest machine machine learning search search engines security security engineering source code systems target text unicode visualization vulnerabilities web security