Trading Devil Final: Backdoor attack via Stock market and Bayesian Optimization

TL;DR

This paper introduces MarketBackFinal 2.0, a backdoor attack on speech-based transformer models using stock market data poisoning, highlighting vulnerabilities in LLM-based speech recognition systems.

Contribution

It presents a novel backdoor attack method leveraging stock market models and acoustic data poisoning to expose security risks in speech-based transformers.

Findings

Demonstrates vulnerability of speech transformers to backdoor attacks

Shows effectiveness of stock market data poisoning in attack implementation

Highlights potential security risks in LLM-based speech systems

Abstract

Since the advent of generative artificial intelligence, every company and researcher has been rushing to develop their own generative models, whether commercial or not. Given the large number of users of these powerful new tools, there is currently no intrinsically verifiable way to explain from the ground up what happens when LLMs (large language models) learn. For example, those based on automatic speech recognition systems, which have to rely on huge and astronomical amounts of data collected from all over the web to produce fast and efficient results, In this article, we develop a backdoor attack called MarketBackFinal 2.0, based on acoustic data poisoning, MarketBackFinal 2.0 is mainly based on modern stock market models. In order to show the possible vulnerabilities of speech-based transformers that may rely on LLMs.