🧰 BaxBench

The items that name it 5

• Qwen3-Coder writes less secure code when told it's for a US persona
  articleas classification.key entities2026-10-21
• Chinese AI Model Writes More Hackable Code When Told It’s for US Contractor
  articleas classification.key entities2026-10-21
• Applying Security by Design at the Point of Execution: How Governed Security Requirements Affect the Security of AI-Generated Code
  paperas classification.key entities2026-10-07
• The Secure Agentic Development Lifecycle (ADLC), Explained | Blog | Endor Labs
  articleas tools2026-09-02

  Benchmark used to measure correctness/security of AI-generated code solutions

• Benchmarking Secure-and-Functional Remediation and How Snyk Agent Fix Lifts Frontier-Model Fix Rates by over 14%
  articleas tools2026-08-18

  Cited benchmark reinforcing that functionally correct code is frequently still insecure

In every workspace you can read 8