13. The 70/30 Rule for AI-Assisted Testing
A useful personal rule is:
70% — Your thinking
Use your own:
- Testing knowledge
- Product understanding
- Domain knowledge
- Logical reasoning
- Risk analysis
- Exploratory testing skills
30% — AI assistance
Use AI for:
- Brainstorming
- Test-case suggestions
- Documentation
- Code assistance
- Explanations
- Test-data ideas
- Reviewing your work
- Summarization
The exact percentage doesn’t have to be 70/30.
The important principle is:
Your thinking should remain in control.
14. Try Testing Without AI Sometimes
This may sound strange in an AI-focused world.
But it is extremely useful.
Take a simple feature.
For example:
Registration Form
Spend 15–20 minutes creating test scenarios without AI.
Think about:
- Functional testing
- Negative testing
- Boundary testing
- Validation
- Security
- Usability
- Compatibility
- Error handling
Then ask AI:
“Review my scenarios and identify gaps.”
Now compare both lists.
This exercise helps you discover whether you are actually improving as a tester.
15. Use AI as a Second Opinion
One of the best uses of AI is getting another perspective.
Imagine you identified a production issue.
You believe the root cause is an API failure.
Before concluding, ask AI:
“Here is the observed behavior and log information. What other possible causes should I investigate?”
AI might suggest:
- Frontend issue
- API issue
- Database issue
- Authentication issue
- Network timeout
- Cache problem
Now you have additional investigation paths.
But you still verify them yourself.
16. Don’t Give AI Sensitive Company Information
This is extremely important.
QA engineers often work with:
- Customer information
- API keys
- Passwords
- Authentication tokens
- Production URLs
- Database credentials
- Internal architecture
- Proprietary source code
- Confidential business requirements
Do not blindly paste sensitive company information into an AI tool.
QA engineers should also follow responsible AI and data-security practices when working with AI tools. The NIST AI Risk Management Framework provides guidance for managing AI-related risks.
Follow your organization’s:
- Security policy
- Data privacy policy
- AI usage policy
- Compliance requirements
When possible, anonymize data.
For example, instead of:
Customer Name: Rahul Sharma
Email: rahul@company.com
Phone: 98765432
17. Don’t Let AI Replace Exploratory Testing
Exploratory testing requires curiosity.
You interact with the application.
You observe unexpected behavior.
You ask:
“What happens if I do this?”
Then:
“What happens if I do this differently?”
And then:
“What happens if I combine these two actions?”
This type of thinking is difficult to completely replace with predefined AI suggestions.
For example:
You are testing an online shopping application.
You add an item to the cart.
Then you change the quantity.
Then you open another browser tab.
Then you change the address.
Then you apply a coupon.
Then you refresh the page.
Then you return to the cart.
Unexpected behavior may appear.
That curiosity is one of the strongest skills a QA engineer can develop.
18. AI Should Not Make You Lazy
There is a difference between:
Working smarter
and
Avoiding thinking
Working smarter means:
“AI can help me complete repetitive work faster, so I can spend more time on risk analysis and quality.”
Avoiding thinking means:
“AI generated the test cases, so my job is finished.”
The first approach makes you stronger.
The second makes you dependent.