AI glossary

Source

A clear explanation of what a source means in AI answers, search results, articles, and suspicious messages.

Edited by H. Omer Aktas

Listen to this page Reads only the article text, not the menu, footer, or right rail.

Ready to read this guide aloud.

Verification rule: A claim without a clear source is only a claim.

Opening answer

A source is where information comes from. In AI, search, news, social media, email, and everyday decisions, the source tells you whether a claim can be checked. A source might be an official help page, government website, company policy page, research paper, product manual, news article, receipt, bank statement, or a person with direct knowledge. The first thing to know is that not every link is a good source, and not every AI answer has a real source. When the topic affects money, health, law, identity, travel, school, or safety, the quality of the source matters more than the confidence of the answer.

Simple summary

  • A source is the origin of a claim or piece of information.
  • Good sources help you verify facts instead of guessing.
  • Official sources are best for rules, prices, accounts, and safety instructions.
  • AI can mention sources incorrectly, so check them yourself.
  • Do not trust a link just because it appears in a polished answer.

Try this prompt

Use this when you want a safer, clearer answer from an AI tool.

Prompt:

For each claim in this answer, tell me what kind of source would verify it. Separate official sources from general articles. Do not invent links. If you are not sure, say what I should check manually.

Prompt:

Look at this message and identify its source. Is it from an official account, a forwarded screenshot, an unknown link, or a claim with no clear source? Give me safe next steps.

Plain-English explanation

When someone says, ‘I read it online,’ the next question is, ‘Where online?’ A source gives that answer. In daily life, a strong source is close to the original information. A bank’s own website is a stronger source for bank policy than a random forum comment. A school office is a stronger source for school rules than a parent group rumor. A medicine label or doctor is a stronger source for health instructions than an AI summary.

AI tools may produce answers with or without links. Even when links appear, they may be too general, outdated, or not support the exact claim. The safest habit is to separate the answer from the source. Use the AI answer to understand the topic, then verify key details where the information actually comes from.

A source can also be missing. If a message says, ‘Your payment failed, click here,’ but the sender is strange and the link does not match the company, the source is weak. Open the official app or website yourself instead of using the link.

How people can use it

  • Ask whether an AI answer has enough support.
  • Check if a message really comes from a company or service.
  • Compare a claim with an official policy page.
  • Look for dates when rules or prices may have changed.
  • Teach family members to ask, ‘Where did this come from?’
  • Decide when to ask a human expert instead of trusting a summary.

Step-by-step guidance

  1. Identify the exact claim you need to verify.
  2. Ask where the information came from.
  3. Prefer original or official sources.
  4. Check whether the source is current.
  5. Check whether the source actually says what the answer claims.
  6. Use a separate trusted route for account or payment issues.
  7. Keep notes of sources for important decisions.

Safety and privacy notes

For suspicious emails and texts, the visible sender name is not enough. Criminals can imitate brands, use look-alike links, or send messages that request personal information. CISA warns that phishing can involve harmful links, attachments, and requests for personal information. Verify account problems from the official app or website, not from the link inside a suspicious message.

Common mistakes to avoid

  • Treating any link as proof.
  • Trusting an AI citation without opening it.
  • Using a screenshot as the only source.
  • Ignoring whether a page is old or from another country.
  • Believing a source supports a claim when it only mentions the topic generally.
  • Clicking the source link inside a suspicious account warning.

Examples

Weak source: a social media screenshot saying a subscription price changed. Better source: the company’s official pricing page or account billing page.

Weak source: an AI answer saying a local rule applies everywhere. Better source: the local government, school, landlord, bank, or official agency that controls the rule.

Weak source: a text message link saying your package failed delivery. Better source: the delivery company’s official app or website typed manually.

Source quality table

Better and weaker sources
Need to checkStronger sourceWeaker source
Account problemOfficial app, official website, known support numberText link or unknown caller
Medical questionDoctor, pharmacist, medicine label, health agencyAI answer or comment thread
Refund ruleCompany policy page or written receiptOld forum post
School policyTeacher, school office, school websiteParent group rumor
AI claimOriginal source that directly supports the claimCitation that only sounds related

What is a source?

A source is the place, document, person, or record where information comes from. In AI and search, sources help readers check whether a claim is real, current, and supported.

What makes a source trustworthy?

A trustworthy source is close to the original information, current, relevant to your situation, and clear about who created it. Official websites, original policies, professional records, and direct documents are usually stronger than screenshots, rumors, or anonymous posts.

Can AI give fake sources?

Yes. AI may produce links, titles, or citations that are wrong, outdated, or not strong enough for the claim. Open important sources yourself and check whether they truly support the answer.

Data and source notes

For scam and phishing checks, use official cybersecurity and consumer resources. CISA’s phishing guidance explains why unknown links and requests for personal information need caution. For AI definitions, NIST’s AI glossary is a stronger source than random summaries.

FAQ

Is Wikipedia a source?

It can be a starting point, but important claims should be checked against original or official sources.

Is an AI answer a source?

Usually no. It is an answer. You still need to know where the information came from.

What if a source link is broken?

Look for the official site or another reliable source. Do not rely on a broken citation.

Can a company blog be a source?

Yes for that company’s own product information, but compare claims with official documentation when needed.

Should I trust sources in ads?

Be careful. Ads are designed to persuade. Check official details before buying.

What is the best source for account warnings?

Your official app, official website, or known support number, not a link in an unexpected message.

Final takeaway

A source is your way to check reality. AI can make information easier to read, but the source decides whether the claim deserves trust. For serious issues, ask where the information came from, check the original place, and avoid links in suspicious messages.