Breaking to Build: A Threat Model of Prompt-Based Attacks for Securing LLMs
It addresses the problem of securing LLMs against adversarial prompt-based attacks for researchers and developers, but it is incremental as it synthesizes existing knowledge rather than introducing new methods.
This paper tackles the security challenges of Large Language Models (LLMs) by presenting a comprehensive literature survey and threat model of prompt-based attacks, which exploit vulnerabilities to cause harm like intellectual property theft and misinformation, aiming to inform the development of more secure LLMs.
The proliferation of Large Language Models (LLMs) has introduced critical security challenges, where adversarial actors can manipulate input prompts to cause significant harm and circumvent safety alignments. These prompt-based attacks exploit vulnerabilities in a model's design, training, and contextual understanding, leading to intellectual property theft, misinformation generation, and erosion of user trust. A systematic understanding of these attack vectors is the foundational step toward developing robust countermeasures. This paper presents a comprehensive literature survey of prompt-based attack methodologies, categorizing them to provide a clear threat model. By detailing the mechanisms and impacts of these exploits, this survey aims to inform the research community's efforts in building the next generation of secure LLMs that are inherently resistant to unauthorized distillation, fine-tuning, and editing.