CWE-1427: Improper Neutralization of Input Used for LLM Prompting
Description
The product uses externally-provided data to build prompts provided to large language models (LLMs), but the way these prompts are constructed causes the LLM to fail to distinguish between user-supplied inputs and developer provided system directives.
When prompts are constructed using externally controllable data, it is often possible to cause an LLM to ignore the original guidance provided by its creators (known as the system prompt) by inserting malicious instructions in plain human language or using bypasses such as special characters or tags. Because LLMs are designed to treat all instructions as legitimate, there is often no way for the model to differentiate between what prompt language is malicious when it performs inference and returns data. Many LLM systems incorporate data from other adjacent products or external data sources like Wikipedia using API calls and retrieval augmented generation (RAG). Any external sources in use that may contain untrusted data should also be considered potentially malicious.
- Abstraction
- BaseA weakness specific enough to point to detection and prevention methods.
- Likelihood of exploit
- Not rated by MITRE
- Status
- Incomplete
Common consequences
- ConfidentialityIntegrityAvailability
Execute Unauthorized Code or Commands; Varies by Context
The consequences are entirely contextual, depending on the system that the model is integrated into. For example, the consequence could include output that would not have been desired by the model designer, such as using racial slurs. On the other hand, if the output is attached to a code interpreter, remote code...
- Confidentiality
Read Application Data
An attacker might be able to extract sensitive information from the model.
- Integrity
Modify Application Data; Execute Unauthorized Code or Commands
The extent to which integrity can be impacted is dependent on the LLM application use case.
- Access Control
Read Application Data; Modify Application Data; Gain Privileges or Assume Identity
The extent to which access control can be impacted is dependent on the LLM application use case.
Potential mitigations
- Architecture and DesignEffectiveness: High
LLM-enabled applications should be designed to ensure proper sanitization of user-controllable input, ensuring that no intentionally misleading or dangerous characters can be included. Additionally, they should be designed in a way that ensures that user-controllable input is identified as untrusted and potentially dangerous.
- ImplementationEffectiveness: Moderate
LLM prompts should be constructed in a way that effectively differentiates between user-supplied input and developer-constructed system prompting to reduce the chance of model confusion at inference-time.
- Architecture and DesignEffectiveness: High
LLM-enabled applications should be designed to ensure proper sanitization of user-controllable input, ensuring that no intentionally misleading or dangerous characters can be included. Additionally, they should be designed in a way that ensures that user-controllable input is identified as untrusted and potentially dangerous.
- Implementation
Ensure that model training includes training examples that avoid leaking secrets and disregard malicious inputs. Train the model to recognize secrets, and label training data appropriately. Note that due to the non-deterministic nature of prompting LLMs, it is necessary to perform testing of the same test case several times in order to ensure that troublesome behavior is not possible. Additionally, testing should be performed each time a new model is used or a model's weights are updated.
- InstallationOperation
During deployment/operation, use components that operate externally to the system to monitor the output and act as a moderator. These components are called different terms, such as supervisors or guardrails.
- System Configuration
During system configuration, the model could be fine-tuned to better control and neutralize potentially dangerous inputs.
Related weaknesses
CAPEC attack patterns that exploit it
CWE 4.20 links no CAPEC attack pattern to this weakness.