Abstract
This disclosure describes AI Code Canaries, a novel approach to source code protection using dynamically generated, context-aware code snippets (canaries) embedded within the codebase to proactively detect unauthorized code leakage. This method addresses limitations of traditional watermarking and other techniques by offering finer-grained detection, increased resilience to code transformations, and immediate leak identification. Unlike traditional watermarking methods, which struggle with fine-grained detection and are vulnerable to code transformations, AI Code Canaries offer granular detection of even small code fragments, increased resilience through dynamic generation, and proactive leak identification.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Van Hoose, Ray, "AI Code Canaries for Proactive Source Code Leak Detection", Technical Disclosure Commons, (December 23, 2024)
https://www.tdcommons.org/dpubs_series/7683