Protect sensitive information
Avoid entering passwords, financial details, confidential business information, private customer data or other sensitive information into an AI system unless you understand how that system handles and protects the data.
Tech Learning Hub · AI Lesson 03
Using AI Safely & Responsibly
Learn practical habits for protecting information, checking AI-generated content, recognising limitations and keeping human judgement at the centre of important decisions.
← Artificial Intelligence Lessons
Published: October 26, 2026
Last updated: October 26, 2026
The basic idea
Safe AI use starts with understanding its limits
Artificial intelligence can help people write, learn, research, organise information, generate ideas and complete many everyday tasks. But useful technology still needs to be used with care.
AI systems do not automatically know whether every answer they produce is true, fair, safe or appropriate. They can make mistakes, misunderstand context and generate information that appears believable even when it is inaccurate.
Responsible AI use means understanding these limitations and making deliberate decisions about what information you provide, what outputs you trust and when a human should make the final decision.
Practical principles
Six habits for responsible AI use
These principles are useful whether you are using AI for school, work, business, research or personal tasks.
Verify important information
AI can generate answers that sound convincing but are incomplete, outdated or incorrect. Important claims should be checked against reliable sources before you act on them.
Keep human judgement involved
AI can assist with research, writing, planning and analysis, but people remain responsible for evaluating the result and deciding whether it is appropriate to use.
Watch for bias
AI systems learn from large amounts of data that may contain social, cultural or historical biases. Their outputs can sometimes reproduce or amplify those patterns.
Use AI honestly
Do not use AI to deliberately deceive people, impersonate others, create fraudulent material or misrepresent AI-generated work when disclosure is required.
Know when AI is not enough
Some decisions require qualified professional advice, direct evidence or human expertise. AI should not automatically replace those sources of judgement.
Privacy & security
Think before you share information with AI
When you type information into an AI service, that information leaves your immediate control and is processed according to the service provider's systems, settings and policies.
Different AI services may handle data differently. Before using an AI tool for sensitive work, understand its privacy controls, data-retention practices and the rules that apply to your organisation or situation.
Important
Do not treat an AI chat like a private notebook
Unless you know that a particular system is approved for the information you are handling, avoid entering confidential or highly sensitive data.
Examples of information that deserves extra care
Passwords, PINs and authentication codes
Bank card or account credentials
Private identification documents
Confidential customer or employee information
Unpublished business information or trade secrets
Private medical, legal or financial documents
Accuracy
Why should AI answers be verified?
A language model is designed to generate useful sequences of language. It is not a guarantee that every sentence it produces has been independently checked against reality.
An AI system may produce incorrect dates, statistics, names, explanations or references. It can also combine accurate and inaccurate information in the same response.
This is especially important when the information could affect someone's money, health, legal position, safety, education or business.
A useful rule
The more important the decision, the more important independent verification becomes.
Verification
A simple way to check an AI response
1
Identify the important factual claims in the AI response.
2
Check those claims against reliable and preferably independent sources.
3
Look for dates, because information can become outdated.
4
Check names, numbers, quotations, references and links carefully.
5
For high-impact decisions, consult an appropriate qualified professional or authoritative source.
Bias & fairness
AI output is not automatically neutral
AI models are developed using data, design choices and human decisions. Those inputs can contain biases or incomplete representations of people, cultures, places and experiences.
As a result, an AI response may sometimes contain stereotypes, unfair assumptions or conclusions that work better for some groups than others.
When AI is used in situations that affect people, such as recruitment, education, lending or access to services, its recommendations should be reviewed carefully rather than treated as automatically objective.
Human responsibility
AI can assist a decision without owning the decision
AI can help summarise options, identify patterns, draft documents and suggest possible approaches. But the person using the system still needs to consider the consequences of acting on its output.
Saying that "the AI recommended it" does not remove human responsibility. People and organisations remain responsible for how AI-generated information is evaluated and used.
AI can help
Brainstorm ideas, organise information, explain concepts, compare possibilities, create drafts and support routine analysis.
Humans should decide
Whether the information is trustworthy, whether using it is fair and appropriate, and what action should ultimately be taken.
Key takeaways
What to remember from this lesson
✓
AI output can sound confident even when the information is incorrect.
✓
Important AI-generated information should be verified before it is trusted or acted upon.
✓
Sensitive or confidential information should be handled carefully when using AI tools.
✓
AI systems can reflect biases found in their training data and design.
✓
Human judgement remains important, especially for high-impact decisions.
✓
Responsible AI use includes honesty, accountability and awareness of possible harm.
Continue learning
Next: AI for Work & Business
Next, explore practical ways artificial intelligence can support everyday work and business tasks while keeping human judgement and responsibility in the process.
Return to AI learning path