01 AI in Software Testing
Where AI genuinely helps QA work today, where it doesn't, and why this path is scoped to using AI with judgment — not prompt engineering, not machine learning.
02 Responsible AI Usage and Human-in-the-Loop QA
What human-in-the-loop concretely means for QA work, and why a review step that exists on paper but not in practice provides no real protection.
03 Reviewing AI Output and Hallucinations
The practiced skill of catching a plausible-sounding but wrong AI output — a fabricated field name, an invented requirement, a confidently wrong root cause — before it enters a test suite.
04 AI-Assisted Test Case Generation
Using AI to draft test cases from a requirement, then applying Boundary Value Analysis and Equivalence Partitioning to systematically evaluate and correct the draft — not accept it wholesale.
05 AI-Assisted Test Data Creation
Using AI to generate realistic test data, evaluated against the same volume, shape, and distribution criteria Test Data Design and Test Data for Performance already established — not accepted because it looks plausible.
06 AI-Assisted API and Automation Authoring
Using AI to draft API test scripts and automation code, reviewed against API Testing's and Automation Testing's own existing standards — not a new framework, an AI-assisted way of authoring within the ones already taught.
07 AI-Assisted Defect Analysis and Exploratory Testing
Using AI to suggest a starting root-cause hypothesis and exploratory testing charters — both treated as a starting point to verify, never a diagnosis or a substitute for genuine human discovery.
08 Testing AI-Driven Features
What a tester validates about AI-generated product content as a distinct feature-testing surface — and why deterministic software defects and AI quality issues are different problem classes needing different evaluation strategies.
09 Prompt Testing and Evaluation
Testing how an AI feature behaves across realistic input variation, and scoring its responses with a structured rubric instead of a pass/fail assertion or ad hoc judgment.
10 Hallucinations, Bias, Safety, Reliability
Testing a shipped AI feature for four related quality defect classes — fabricated claims, unfair treatment across phrasing, unsafe scope violations, and inconsistent or ungrounded answers.
11 AI Governance for QA
What a QA team needs in place before adopting AI tools at scale — an explicit usage policy, defined review requirements, and clear accountability for AI-assisted work products.
12 AI Security and Privacy Awareness
What data is unsafe to send to an external AI tool, and prompt-injection-style risks specific to AI-assisted testing workflows — scoped to QA-relevant awareness, not a security-specialist treatment.
13 Human Review Workflows and AI QA
Bringing together every review standard this path has taught into one operational workflow — routing each AI-assisted artifact type to its correct, specific review before it's trusted.
14 AtlasBank AI Support Assistant Validation
Combining prompt testing, rubric evaluation, and hallucination/bias/safety/reliability testing from Section 3 against the real, integrated AtlasBank AI Support Assistant.
15 AI-Assisted End-to-End QA Workflow
Combining test case generation, test data creation, automation authoring, and defect analysis from Section 2 into one AI-accelerated workflow, routed through Section 4's unified review process.
16 Common Mistakes in AI for QA
Six recurring mistake patterns spanning review discipline, hallucination trust, test design coverage, AI-feature testing method, data safety, and workflow consistency — each traced to a real defect earlier in this path.
17 Capstone: AI Support Assistant Verification
The complete AI-for-QA lifecycle applied to TestAtlas's first AI-native feature — building the AI Support Assistant's own test suite with AI assistance, then testing the Assistant itself, closing with what human review caught that AI-assisted testing alone would have missed.
Section 1 — Review
A recap of Foundations of AI for QA — where AI helps and doesn't, genuine human-in-the-loop review, and recognizing hallucinations — plus a Knowledge Check.
Section 1 — Solutions
Answers and reasoning for Section 1's Knowledge Check — applying the drafting-vs-judgment distinction, genuine review, and hallucination recognition to five realistic scenarios.
Section 2 — Review
A recap of AI-Assisted Testing Techniques — test case generation, test data creation, API and automation authoring, and defect analysis/exploratory testing — plus a Knowledge Check.
Section 2 — Solutions
Answers and reasoning for Section 2's Knowledge Check — applying BVA review, test data evaluation, automation-quality review, and defect-hypothesis verification to five realistic scenarios.
Section 3 — Review
A recap of Testing AI-Driven Features — deterministic vs. AI-quality defects, prompt testing and rubric evaluation, and hallucination/bias/safety/reliability testing — plus a Knowledge Check.
Section 3 — Solutions
Answers and reasoning for Section 3's Knowledge Check — applying the deterministic-vs-AI-quality distinction, prompt testing, and hallucination/bias/safety/reliability testing to five realistic scenarios.
Section 4 — Review
A recap of AI Governance and Security — governance policy structure, data safety and prompt injection awareness, and the unified human review workflow — plus a Knowledge Check.
Section 4 — Solutions
Answers and reasoning for Section 4's Knowledge Check — applying governance policy structure, data safety, and the unified review workflow to five realistic scenarios.