Prompt Security Introduces Vulnerable Code Scanner for AI-Generated Code

Prompt Security has launched the beta version of its Vulnerable Code Scanner, a tool designed to identify potential risks in AI-generated code before they can affect enterprises, announced in a press release. This new feature automatically scans AI-generated code to prevent developers from using hazardous code suggestions from generative AI applications.

The Vulnerable Code Scanner complements Prompt Security's existing capabilities, which aim to prevent the exfiltration of sensitive information such as organizational secrets, personally identifiable information (PII), and intellectual property (IP). By scanning AI-generated responses, the tool blocks malicious code before it reaches developers, providing a comprehensive protection layer between developers, large language models (LLMs), and codebases.

The scanner supports various AI code assistants, including ChatGPT, Microsoft Copilot, and others, and works with nearly 30 programming languages. Prompt Security also announced its support for Cursor, an AI code assistant, which will now benefit from automatic redaction of sensitive information and other security features.

Prompt Security plans a gradual rollout for additional AI code assistants like GitHub Copilot and Tabnine, further expanding its security offerings for developers using AI tools.