Using ChatGPT to Generate Test Cases



Using ChatGPT to Generate Test Cases

ChatGPT can significantly speed up test design by generating functional, negative, boundary, security, and exploratory test scenarios from requirements or user stories. However, the generated test cases should always be reviewed by a QA engineer to ensure they match the application’s business rules.

Why Use ChatGPT for Test Case Generation?

Benefits

  • Saves time creating test cases
  • Improves test coverage
  • Generates positive and negative scenarios
  • Suggests edge cases
  • Creates boundary value tests
  • Helps identify missing requirements
  • Generates API and database test scenarios
  • Produces BDD scenarios
  • Assists junior QA engineers

Step 1: Provide Clear Requirements

Instead of asking:

Generate login test cases.

Provide detailed requirements.

Example Requirement

Application: Online Shopping Portal

Login Requirements:

• Username is mandatory.
• Password is mandatory.
• Username must be a valid email address.
• Password must contain at least 8 characters.
• After 5 failed login attempts, the account is locked for 30 minutes.
• Forgot Password link is available.
• Successful login redirects users to the dashboard.

The more context you provide, the better the output.

Step 2: Use an Effective Prompt

Example Prompt

You are a Senior QA Engineer.

Generate comprehensive test cases for the following login functionality.

Include:

• Positive test cases
• Negative test cases
• Boundary value analysis
• Equivalence partitioning
• Security test cases
• UI validation
• Usability test cases

Present the output in a table with:

Test Case ID
Scenario
Preconditions
Test Steps
Expected Result
Priority

Step 3: Review the Generated Test Cases

Example output:

ID Test Scenario Expected Result
TC001 Login with valid username and password Dashboard displayed
TC002 Invalid password Error message displayed
TC003 Invalid email format Validation message shown
TC004 Blank username Username required
TC005 Blank password Password required
TC006 Both fields blank Mandatory field validation
TC007 Password with 7 characters Validation error
TC008 Five failed login attempts Account locked
TC009 Forgot Password link Reset page opens
TC010 SQL Injection attempt Login blocked
TC011 XSS payload in username Input sanitized
TC012 Session timeout Redirect to login page

Generate Boundary Value Test Cases

Prompt:

Generate boundary value test cases for the Password field.

Rules:
Minimum length = 8
Maximum length = 20
Example output:
Input Expected Result
7 characters Rejected
8 characters Accepted
9 characters Accepted
19 characters Accepted
20 characters Accepted
21 characters Rejected

Generate API Test Cases

Prompt:

Generate REST API test cases for the Login API.

Request:

POST /login

Fields:
email
password

Include:
Positive
Negative
Security
Performance
Authorization
Example scenarios:
  • Valid credentials
  • Invalid password
  • Missing email
  • Missing password
  • Invalid JSON
  • Invalid Content-Type
  • SQL Injection payload
  • XSS payload
  • Rate limiting
  • Large request body
  • Response time validation
  • Invalid token

Generate Exploratory Testing Ideas

Prompt:

Suggest exploratory testing scenarios for an online banking application.
Example output:
  • Navigate rapidly between pages
  • Refresh during money transfer
  • Open multiple browser tabs
  • Use browser back button
  • Interrupt network connectivity
  • Test session timeout
  • Upload unsupported files
  • Test concurrent user actions

Generate Security Test Cases

Prompt:

Generate security test cases for a login page.

Include:

SQL Injection
XSS
CSRF
Session Management
Authentication
Authorization
Example scenarios:
  • SQL injection in username
  • XSS payload in password
  • Brute-force login attempts
  • Weak password policy
  • Cookie tampering
  • Session fixation
  • Logout and browser back button
  • Unauthorized URL access

Generate Test Data

Prompt:

Generate 50 valid and invalid email addresses for testing.
ChatGPT can also generate:
  • Phone numbers
  • Credit card formats (non-real)
  • ZIP/postal codes
  • Names
  • Addresses
  • Dates
  • Random IDs

Generate BDD Scenarios

Prompt:

Generate Gherkin scenarios for Login functionality.
Example:
Feature: Login
Scenario: Successful Login
Given the user is on the Login page
When the user enters valid credentials
And clicks Login
Then the Dashboard should be displayed

Generate Automation Scripts

Prompt:

Generate Playwright TypeScript automation for Login.
Example:
test('Successful Login', async ({ page }) => {
    await page.goto('https://example.com');

    await page.fill('#email', 'user@example.com');
    await page.fill('#password', 'Password123');

    await page.click('button[type="submit"]');

    await expect(page).toHaveURL(/dashboard/);
});

Best Prompt Templates

Template 1 – Functional Testing

Generate detailed functional test cases for:

<Application Name>

Include:
Positive
Negative
Boundary
Validation
Business Rules

Output as a table.

Template 2 – API Testing

Generate REST API test cases.

API:
<Request>

Include:
Status Codes
Headers
Authentication
Security
Performance
Invalid Requests

Template 3 – Automation

Generate Playwright TypeScript automation scripts for:

<Application Feature>

Use:
Page Object Model
Assertions
Best Practices

Template 4 – Bug Report

Create a professional bug report.

Include:
Title
Steps
Expected
Actual
Environment
Severity
Priority

Best Practices

  • Provide complete requirements and business rules.
  • Specify the desired output format (table, BDD, checklist, etc.).
  • Ask for positive, negative, boundary, and edge-case scenarios.
  • Request security, performance, and usability test ideas where applicable.
  • Review AI-generated test cases for accuracy and completeness.
  • Update the generated test cases to reflect organization-specific requirements.

Common Mistakes to Avoid

  • Using vague prompts with little context.
  • Assuming AI-generated test cases are complete.
  • Ignoring business-specific validation rules.
  • Not validating edge cases manually.
  • Skipping reviews before adding test cases to your test repository.

Example Workflow

Requirements
      │
      ▼
Create a Detailed Prompt
      │
      ▼
ChatGPT Generates Test Cases
      │
      ▼
QA Reviews and Refines
      │
      ▼
Add Missing Business Scenarios
      │
      ▼
Convert to Automation (Playwright/Selenium)
      │
      ▼
Execute Tests
      │
      ▼
Analyze Results

Related Posts

Tags: ,
Leave a comment

Your email address will not be published. Required fields are marked *

Subscribe now

Receive weekly newsletter with educational materials, new courses, most popular posts, popular books and much more!