Despite the widespread adoption of AI tools in software development, trust in AI-generated code remains a significant concern. A recent Stack Overflow Developer survey indicates that nearly 40% of developers lack confidence in AI-generated code, and only 2.3% of professional developers highly trust its accuracy. This trust gap currently hinders the full adoption of autonomous AI agents, as developers seek more transparency and understanding of AI-driven processes.
In order to proactively address this concern and in response to early access feature testing and feedback, we’re excited to be able to announce the general availability of a new feature called Test Review.
Introducing Test Review: Bridging the Gap
Diffblue Cover’s latest Test Review feature addresses these concerns by providing developers with a comprehensive way to review and validate AI-generated tests before they get added to your codebase.
Test Review gives developers the ability to edit, analyse and manage each of the test cases that Diffblue Cover autonomously writes for you, as soon as those tests are available.
This feature allows for:
- Enhanced Transparency: Developers can inspect tests written, gaining insights into what is being tested, assertion detail and building trust through hands-on verification.
- Flexibility and Control: Users can choose between using the fully autonomous agent mode or the Test Review interactive mode, tailoring their approach to suit specific project needs.
- Ownership and Accountability: By actively reviewing and refining tests, developers can take greater ownership of the code, gain deeper understanding, and remain accountable for both AI-generated code and overall code quality.
Why Test Review Matters
Balancing productivity with reliability is more important than ever. Test Review offers a unique solution by combining the efficiency of AI with the transparency some developers may need to inspect, understand and fully trust AI generated code. This approach ensures that developers/testers benefit from fully autonomous unit test generation at scale and combines it with optional human oversight.
Key Benefits of Diffblue Cover’s Test Review feature
- Autonomous Test Generation: Diffblue Cover generates tests 4x faster than GitHub Copilot and 10x as many test cases in the same amount of time as a Copilot-assisted developer.
- Human Oversight: Developers maintain control and insight into the testing process.
- Flexibility: Developers can choose between fully autonomous agent mode and assistive Test Review modes based on project requirements.
- Trust and Transparency: An interactive way to build confidence in AI-generated code through review and verification.
- Enhanced Ownership: Developers take greater accountability for code quality compared to other AI-driven coding tools.
Wrapping up
Test Review is now available for all Diffblue Cover users. This feature is designed to accelerate the adoption of AI-powered testing by providing a stepping stone for developers who want to leverage AI without sacrificing their ability to understand and be accountable for the code generated.
At Diffblue, we’re committed to empowering developers with tools that enhance productivity while ensuring the highest code quality standards. With Test Review, we’re adding a flexible option for collaboration between humans and AI in software testing. Whether you prefer full autonomous agent mode or a more interactive Test Review experience, Diffblue Cover adapts to fit your workflow.
We want to build trust in AI-generated code by giving developers the ability to scrutinize and verify the quality of AI generated unit test cases. We can’t wait to hear your feedback on this feature and see what it enables you to build next!
Learn more
Learn more about Test Review in docs
New to Diffblue Cover? Try Diffblue Cover for free
Have questions/ feedback? The Diffblue Community is here to help
