Prompt Security adds code sanitization, data leak prevention for GitHub Copilot

Prompt Security has announced an upgrade to its security and governance platform for GitHub Copilot and other AI coding assistants. The enhancements to the platform, also named Prompt Security, include stronger capabilities for real-time code sanitization and redaction to prevent leaks of sensitive data, the company said.

The update, announced January 20, will be released gradually throughout the first quarter of 2024. The update brings added support for GitHub Copilot Free, Amazon Q, Tabnine, Blackbox, GitLab Duo, Supermaven, and Sourcegraph Cody. Prompt Security also now scans for vulnerable output in the AI code assistants’ responses, blocking generated code that identifies as potentially hazardous or containing vulnerabilities.

The platform’s data leak prevention capabilities prevent exfiltration of secrets, personally identifiable information, and intellectual property while providing visibility and monitoring capabilities. The platform now also will inventory all of the AI tools in use, including the developers who are using the enterprise release of GitHub Copilot versus the free version.