AI-Assisted QA
Scrum Manager's guide IA aplicada al trabajo ágil [AI Applied to Agile Work, in Spanish] describes AI-assisted QA as one of the AI applications with the greatest practical return in agile teams, but also as one of the most prone to happy path bias: the tendency to test only the ideal flow and leave error scenarios uncovered.
Test case generation
The difference between quantity and coverage
Having many test cases does not guarantee good coverage. A hundred cases that test variations of the ideal flow can leave uncovered a single error scenario that causes 80% of production incidents. Meaningful coverage is measured by how many different risks are covered, not by how many cases exist.
Framework for generating test cases with AI
- Start from the acceptance criteria. Each criterion should have at least one case that verifies it.
- Add negative scenarios. For each flow, identify what could fail and how the system should behave.
- Include edge cases: boundary values, empty lists, extreme formats, inconsistent states.
- Link to known risks. If there are areas of the system that have historically failed, ensure specific coverage.
- Prioritize by impact: probability of failure × impact of failure × cost of late detection.
- Review looking for gaps: are there criteria without cases? Are there failure types without coverage?
The happy path bias
The most damaging defects tend to appear in: invalid or malformed inputs, error and recovery states, capacity and throughput limits, insufficient or excessive permissions, concurrency and race conditions, and interruptions and intermediate states. A test plan that only covers the ideal flow is incomplete by definition.
AI tends to generate cases for the happy path because it is the most heavily represented pattern in its training data. The QA engineer must explicitly request negative scenarios and edge cases.
Risk-Focused Assisted Code Review
AI can assist in code review by identifying known risk patterns: common security vulnerabilities, deprecated functions, incorrect error handling, duplicated logic, and inconsistencies with the project's style.
Assisted review follows three guiding questions:
- Does this code do what the spec or the user story asked for?
- Are there code paths that could fail silently?
- Are there any obvious security risks?
Traceability to requirements
Every test case should be traceable to an acceptance criterion, an identified risk, or a specific requirement. Cases without traceability are candidates for being redundant or irrelevant. Traceability also makes it possible to identify gaps: if an acceptance criterion has no associated case, there is an uncovered risk.
AI-assisted QA and the reinforced DoD
AI-assisted QA is integrated into the reinforced DoD for teams working with AI in two dimensions:
- As a tool for generating tests that validate agent-generated code.
- As a layer of AI supervising AI that automatically reviews the code before human review.
Human review remains the final link in the quality chain: AI-assisted QA reduces the time it takes to detect mechanical defects, but validating business logic requires human judgment.
Common mistake
Assuming that more AI-generated cases means better coverage. AI can generate hundreds of cases in minutes. Without selection and prioritization criteria, the result is a voluminous test plan that is biased toward the happy path and full of redundant cases. The quality of AI-assisted QA depends on the quality of the instructions given to the AI (which risks to cover, which negative scenarios to include), not on the number of cases generated.
Resources
🏦 IA aplicada al trabajo ágil [AI Applied to Agile Work, in Spanish]Skill Arena · Scrum Manager
🏦 Scrum en equipos con IA [Scrum in Teams with AI, in Spanish]Skill Arena · Scrum Manager
📊 Guía didáctica Scrum en equipos con IA [Teaching Guide: Scrum in Teams with AI, in Spanish]Resources · Scrum Manager
See also
Want to advance in agility? You can search for course and exam calls or go at your own pace by becoming a member of the Agile Club. This membership includes exclusive resources, e-learning classrooms, and access to Skill Arena: a space to practice and measure your agile skills at your own pace.