Benchmark RadarAI BENCHMARK PROFILE
HybridCodeAuthorship
HybridCodeAuthorship is a benchmark of Python files with interleaved human- and AI-authored lines, built from CodeSearchNet, for developing and evaluating AI-generated code detection algorithms at line- and chunk-level.
- Released
- 2026-06-10
- Readiness
- Paper only
- Primary field
- General AI
Why it matters
Addresses the need for realistic benchmarks for detecting AI-generated code in industry codebases, which is important for risk management and productivity analysis. Provides a challenging testbed for detection algorithms.
Motivation
Thanks to the rapid adoption of AI code assistants powered by large language models (LLMs), industry codebases are, increasingly, a hybrid of AI- and human-authored code.
Primary resources
Benchmark Radar records only publicly supported details and links back to primary sources for verification.