How to Know What's True on the Internet

Slow down: your instincts are being exploited

The first thing to understand is that viral misinformation is engineered to bypass your critical thinking. It is designed to make you feel something: outrage, fear, vindication, before you have time to think. The emotional hit comes first. The reasoning comes later, if at all.

This is not a character flaw. It is how human brains work. Emotional responses are fast; analytical thinking is slow. Bad actors know this and exploit it deliberately.

The single most effective thing you can do is pause. Before sharing, before reacting, before even finishing the article, ask yourself: why does this make me feel so strongly? Strong emotional reactions to news content are often a signal, not a confirmation.

Check who is actually saying it

Misinformation rarely announces itself. It usually arrives dressed as legitimate reporting, with a credible-sounding domain name, a professional layout, and confident language.

Look at the source. Not just the name, but when the site was created, who owns it, what else it publishes, and whether the URL looks slightly off. Misinformation sites often mimic real outlets with subtle changes.

A useful habit: before trusting a source you have not seen before, search the outlet name plus "bias" or "reliability." Sites like Media Bias/Fact Check maintain records of outlet credibility.

Find the original source

Most viral misinformation is a copy of a copy of a copy. A tweet screenshots a headline. A Facebook post shares the tweet. A blog embeds the Facebook post. By the time it reaches you, the original context, and often the original meaning, has been stripped away.

Always try to find where the claim originated. If someone says "scientists have proven X," find the actual study. If a politician allegedly said something, find the original video or transcript. If a statistic is cited, find where that number actually came from.

This one step eliminates a significant portion of misinformation, because many viral claims are misquotes of real things, statistics taken out of context, or summaries that subtly changed the meaning of the original.

Reverse image search everything

Images lie. Not because they are always fake, though deepfakes are increasingly common, but because real images from one event get reused to illustrate completely different events.

A photo of a protest from 2019 gets shared as if it happened yesterday. A natural disaster image from one country gets attributed to another. An out-of-context screenshot of a crowd gets used to claim attendance at an event.

Reverse image search is your fastest defense. Right-click any image and search Google Images, or use TinEye. You will often find the original context within seconds, including when the image was first published and where it actually came from.

Use it especially for: breaking-news images, casualty claims, screenshots of alleged documents or messages, and before-and-after comparison photos.

Read past the headline

Headlines are written to get clicks, not to convey accurate information. This is true of legitimate outlets as much as dubious ones. The economic incentive to write a provocative headline is universal.

The actual article often tells a more nuanced story. A headline that says "Scientists Warn Coffee Causes Cancer" may lead to an article about a weak correlation in mice at extremely high doses. The headline can be technically defensible and still deeply misleading.

Never share based on a headline alone. Read the full article. Then ask whether the headline actually reflects what the article says.

Look for corroboration, not just confirmation

If something important happened, multiple independent outlets will be reporting it. Not outlets that are all citing the same wire service, but genuinely independent sources with different reporters, different sources, and different angles.

If a claim is only appearing in one place, or only on outlets that share a clear ideological alignment, be skeptical. Real events generate coverage from across the spectrum. The absence of mainstream coverage for a major claim is itself a data point.

This does not mean mainstream outlets are always right and alternative ones are always wrong. It means the burden of proof is higher when a claim is isolated.

Real images can still mislead

One of the more sophisticated forms of misinformation does not use fake images at all. It uses real ones, framed dishonestly.

A photo of a large crowd is real, but the caption says it is for a different event. A video of genuine violence is real, but the location and date are wrong. A statistic is accurate, but the comparison that makes it look alarming uses a misleading baseline.

Ask yourself: even if this image or number is real, does the claimed context actually match? Real facts assembled dishonestly are still misinformation.

Pay special attention to screenshots

Screenshots are the format of choice for misinformation in 2026 because they are hard to quickly verify and feel inherently authentic: a screenshot of a tweet, a text message, or a news article.

Screenshots can be fabricated in minutes. Text can be edited. Usernames can be spoofed. Dates can be changed.

If a screenshot is making a serious claim, find the original. If it is allegedly a tweet, find the tweet. If it is a news article, find the article. If the original does not exist or says something different, you have your answer.

Know your own biases

This is the uncomfortable one. Research consistently shows that people are significantly worse at detecting misinformation when it confirms what they already believe. We apply critical thinking selectively: skeptically to claims that challenge us, generously to claims that validate us.

This means the misinformation that reaches you is often targeted to your existing worldview. It is more likely to look credible to you than to someone with different beliefs. It is more likely to make you feel righteous rather than suspicious.

The antidote is not to distrust everything you agree with. It is to apply the same verification standards to claims you like as to claims you dislike. If you would demand a source for a claim from the other side, demand one for your own side too.

Use tools that do the heavy lifting

You do not have to do all of this manually every time. A growing set of tools can help.

For claims and articles: fact-checking extensions can verify claims against real sources in seconds, identifying whether a claim is supported, contradicted, or unverified, and showing which sources the verdict is based on.

For images: Google Reverse Image Search and TinEye can trace where an image first appeared and in what context.

For source credibility: Media Bias/Fact Check and NewsGuard maintain ratings for thousands of outlets based on their track record for accuracy.

For websites: The Wayback Machine can show what a website looked like in the past, which helps when a newly credible-looking outlet has a suspicious history.

The goal is not to spend ten minutes verifying every piece of content you see. It is to build fast habits that catch obvious problems before they spread.