AI for Software Testing

Begin by exploring what generative AI actually does and how it applies to software testing in practice. You will compare outputs from different AI tools, observe how confident-sounding responses can still be incomplete or incorrect, and examine how AI changes testing work—not by replacing tester judgment, but by accelerating certain activities. This module establishes AI as a collaborator whose output must be evaluated, challenged, and corrected.

Jump directly into test development by asking AI to generate test cases with minimal context. Attempt to execute those tests and discover where AI output helps, where it breaks down, and where assumptions have been silently introduced. By comparing results from different AI tools, you will learn to recognize the difference between tests that appear reasonable and tests that are actually executable and grounded in observed behavior.

AI was able to generate tests from little more than a snapshot of a UI. In this module, you work in the opposite direction using tests and observed behavior to infer requirements and business rules. You will examine how AI fills gaps with assumptions, learn to separate confirmed behavior from speculation, and use requirement taxonomies to organize and correct AI-generated statements. The focus is on requirements as testable descriptions, not authoritative truths.

Real test cases require more than placeholder phrases such as "enter a valid value." They require meaningful data choices that support coverage. In this module, you use equivalence partitioning and boundary value analysis as thinking tools to structure input domains and identify important conditions. AI is used to suggest partitions and boundaries, which you then evaluate for relevance, assumptions, and completeness before selecting representative data values.

What does it actually mean for a test case to be executable? This module focuses on the information required for another tester to run a test without interpretation. Use AI to refine test cases to include clear steps, concrete data, and observable outcomes. AI is used to suggest refinements, then you evaluate whether those refinements achieve clarity or introduce new problems.

User stories describe changes to a system, and acceptance criteria define the conditions under which those changes are considered complete. In this module, you examine how acceptance criteria, scenarios, and test cases relate in practice. You will explore the many-to-many relationship between them and use AI to draft and refine acceptance criteria while ensuring that scenarios remain grounded in realistic system behavior.

Although tests for single pages and functions are a good start and can catch a lot of bugs, eventually we'll need to test larger units of functionality. These can be abstracted as different states and visualized graphically. Learn state-based thinking as a way to model sequences, transitions, and cumulative effects. Use AI to help sketch state diagrams and identify valid transitions, then reason about functional coverage in terms of states, transitions, and invariants rather than test counts.

Functional correctness is not enough. Systems must also meet quality expectations such as usability, accessibility, performance, and security. Use AI to evaluate nonfunctional requirements and to convert these from vague aspirations into solid testable objectives. Design tests that ensure systems meet quality criteria essential for user satisfaction and system reliability.

Individual tests do not add up to a test strategy. Learn how AI can help you to make deliberate testing choices aligned with business goals and risk. Use a Test Strategy canvas to outline and plan a comprehensive testing strategy. See how AI can draft inputs and challenge assumptions, while responsibility for prioritization and trade-offs remains with the tester.

Bug reports are both technical findings and communication artifacts. Practice using AI to analyze bug reports to distinguish real failures from expected behavior or misunderstandings, assess impact and risk, and communicate findings effectively to different audiences. Although AI is used to assist with summarization and categorization, learn to evaluate where this help improves clarity and where it distorts meaning or priority.

Automating test execution promises to make it easier and cheaper to run large test suites more frequently—a practice essential to modern agile development and DevOps practices. Explore the different kinds of automated tests. See how AI can generate automated tests for different levels and architectures. Use AI to evaluate existing systems' suitability for automation and to check and improve existing automation suites.

So what does all of this mean? Now it all comes together. You'll assess where AI can add the most value in your own practice, plan small experiments, and design an ethical framework for responsible adoption. The focus is on practical next steps—how to pilot, measure, and lead change as an AI-empowered tester.