Trust, But Mutate: How to Use Mutation Testing to Assess AI-Generated Tests

​Trust, But Mutate: How to Use Mutation Testing to Assess AI-Generated Tests

​AI tools like GitHub Copilot and ChatGPT are generating test code faster than ever. But how do you know if those tests are actually catching bugs, or just providing a false sense of security?

​Join Cedar Quality Lab for an in-depth technical session on how to objectively measure the quality of your automated tests in the AI era.

​When you rely on AI to write your tests, traditional metrics like code coverage aren't enough. Mutation testing offers a powerful, code-centric way to "test your tests" by deliberately introducing faults (mutations) into your application code and seeing if your test suite actually catches them. If the tests pass despite the broken code, the AI-generated tests have failed you.

​What you will learn:

• ​The core principles of mutation testing and why standard code coverage falls short.

• ​How to identify "fake certainty" and shallow assertions in AI-generated test code.

• ​Practical strategies for integrating mutation testing into your quality engineering workflow to audit AI output.

• ​How to ensure your automation suites actually deliver fast, valuable, and reliable feedback.

​About the Speaker Bas Dijkstra is an independent test automation consultant and trainer with nearly two decades of experience. Based in The Netherlands, Bas is dedicated to helping engineering teams get faster, more valuable feedback from their test automation. As the creator of RestAssured.Net and a seasoned speaker, he helps organizations cut through the automation noise and improve their testing efforts through the smart application of automation tools.