If a unit test fails when the code under test hasn't functionally changed then it means that your unit tests are flaky.