Q1 of 21 · AI for testing

How do you use AI to help generate test cases from a user story?

AI for testingJuniorai-for-testingtest-generationuser-storiesai-toolstest-design

Short answer

Short answer: Provide the user story and acceptance criteria to a language model and ask it to list test scenarios covering happy path, negatives, and edge cases. Treat the output as a draft checklist — review every item, cut irrelevant cases, and add any domain-specific coverage the model missed.

Detail

AI accelerates the first pass — turning an acceptance criterion into 15–20 candidate scenarios in seconds. This is useful when you're time-pressured or working in an unfamiliar domain.

The limitations are real. AI does not know your application's actual behaviour, so "upload fails gracefully when the server is down" appears as plausible as "upload succeeds" — but the first requires you to know your error-handling code. It also misses undocumented business rules that are not in the story text.

Workflow: use the AI output as a structured brainstorm, not a finished test plan. Compare it against the acceptance criteria, add domain knowledge the model lacked, and cut cases that don't reflect real risk. See AI-generated test plans and Requirements to test cases for deeper workflows.

// WHAT INTERVIEWERS LOOK FOR

Draft-and-review workflow, not blind acceptance. Awareness that AI misses undocumented business rules. Knowing it speeds up breadth discovery, not correctness judgement.

// Related questions

How do you build a test strategy when there are no written requirements?

Test design