This is a draft list of important vulnerability types for Artificial Intelligence (AI) applicaitons build on Large Language Models (LLMs)
Description:
Bypassing filters or manipulating the LLM using carefully crafted prompts that make the model ignore previous instructions or perform unintended actions.
Description:
Accidentally revealing sensitive information, proprietary algorithms, or other confidential details through the LLM's responses.
Description:
Failing to properly isolate LLMs when they have access to external resources or sensitive systems, allowing for potential exploitation and unauthorized access.
Description:
Exploiting LLMs to execute malicious code, commands, or actions on the underlying system through natural language prompts.
Description:
Exploiting LLMs to perform unintended requests or access restricted resources, such as internal services, APIs, or data stores.
Description:
Excessive dependence on LLM-generated content without human oversight can result in harmful consequences.
Description:
Failing to ensure that the LLM's objectives and behavior align with the intended use case, leading to undesired consequences or vulnerabilities.
Description:
Not properly implementing access controls or authentication, allowing unauthorized users to interact with the LLM and potentially exploit vulnerabilities.
Description:
Exposing error messages or debugging information that could reveal sensitive information, system details, or potential attack vectors.
Description:
Maliciously manipulating training data or fine-tuning procedures to introduce vulnerabilities or backdoors into the LLM.