# Misinformation vs disinformation: The difference is intent

> A fake evacuation notice forwarded innocently by a neighbor and the same notice fabricated by an agency are two different problems distinguished by intent. What separates misinformation from disinformation, how false claims travel in 2026, and 5 checks to sort them.

Source: https://informed.now/blog/misinformation-vs-disinformation

By Shrestha Saha · October 23, 2026 · News Curation

**Summary:** Misinformation is false information passed along by someone who believes it. Disinformation is false information passed along by someone who knows it's false and wants you to believe it. The claim can be word for word the same, and the difference is intent, which decides how the false claim behaves, who is responsible for it, and what to do about it once you've checked it.

### One false claim, two different problems

The same false sentence can be two different problems. A neighbor who forwards a fake evacuation notice because she thinks it's real has made a mistake, and a correction ends it. Someone who wrote that notice on purpose, to empty the town, has done something deliberate, and a correction doesn't stop him. It only shows him which part of the lie failed to convince. The words on the screen are identical, and everything that matters about them is not.

That is why the two terms exist, and why using them interchangeably hides the thing you most need to know about a false claim. Misinformation is a problem of knowledge, and more knowledge solves it. Disinformation is a problem of motive. The person spreading it already knows the facts, so showing them the facts changes nothing.

### Where the line sits

The United Nations draws the line the way most researchers do. In its [guidance](https://www.un.org/en/countering-disinformation) on countering disinformation, misinformation is inaccurate information spread by accident, while disinformation is inaccurate information that "intends to deceive and is spread in order to do serious harm." The UN also notes that there is no universally accepted definition of disinformation, which is worth remembering whenever a government or a platform defines it to suit itself.

Three features follow from the definitions. Misinformation needs a believer at every link in the chain, because once someone stops believing it, they stop sharing it. Disinformation needs a believer only at the end, since the people who start it never believed it. And a single false claim can change category as it travels: It leaves a troll farm as disinformation and arrives in your family group chat as misinformation, carried by an uncle who means no harm at all.

### Why intent changes the response

Misinformation yields to a correction, because the person sharing it wants to be right and will usually accept being shown otherwise. Disinformation doesn't yield to a correction, because its author already knows. What works against disinformation is exposure, which means naming who made the claim and why, so that the next claim from the same source arrives with its credibility already spent.

Intent changes who answers for it. When the U.S. Justice Department moved against a Russian campaign in September 2024, it seized 32 internet domains, one of them washingtonpost.pm, built to pass as The Washington Post. According to the department's [announcement](https://www.justice.gov/usao-edpa/pr/justice-department-disrupts-covert-russian-government-sponsored-foreign-malign), three companies ran the operation, known as Doppelganger, under the direction of a first deputy chief of staff in the Russian Presidential Administration, with the stated aim of reducing international support for Ukraine and influencing voters in U.S. and foreign elections. Nobody prosecutes a neighbor for a mistaken forward message. However, governments do prosecute the deliberate version, and the charge was never that the sites were wrong, but that they pretended to be someone else on purpose.

And intent changes how you treat the person in front of you. Someone who shares misinformation is a reader who was fooled, and the useful response is to show them the better source without contempt, since contempt is what turns a mistaken reader into a committed one. Someone who shares disinformation is not confused, and the useful response is to stop giving them an audience.

### How false claims travel in 2026

False claims have never had a wider pipe, and the audience knows it. In the Reuters Institute's 2026 Digital News [Report](https://reutersinstitute.politics.ox.ac.uk/digital-news-report/2026/dnr-executive-summary), drawn from 97,520 people across 48 markets, 62% said they were concerned about what is real and what is fake online, up four points in a year, with concern rising in every Western European market surveyed. In the United States, [Pew](https://www.pewresearch.org/short-reads/2025/10/29/many-americans-say-they-often-come-across-inaccurate-news-and-have-a-hard-time-knowing-whats-true/) found in an August 2025 survey of 5,153 adults that nine in ten at least sometimes come across news they think is inaccurate, 42% say it happens often or extremely often, and 51% find it difficult to tell what's true. The World Economic Forum's Global Risks [Report](https://reports.weforum.org/docs/WEF_Global_Risks_Report_2026.pdf) for 2026, built on the views of more than 1,300 experts, ranks misinformation and disinformation second among all global risks for the next two years, behind only geoeconomic confrontation.

Who moves a false claim is the part most people get wrong. The largest study of the question, published in Science in 2018 by three MIT researchers, [tracked](https://news.mit.edu/2018/study-twitter-false-news-travels-faster-true-stories-0308) about 126,000 story cascades shared by some 3 million people on Twitter between 2006 and 2017. False stories were 70% more likely to be retweeted than true ones, and true stories took about six times as long to reach 1,500 people. When the researchers removed every bot from the data, the gap stayed, which means people, not machines, did the spreading. The study is eight years old and the platforms have changed since, but no later study we know of has overturned its central finding: A false claim travels on willing human hands.

Since 2025 there has been a new source of misinformation that has no intent at all. In October 2025, 22 public broadcasters in 18 countries checked more than 3,000 answers that most AI platforms gave about the news, and [found](https://communication.vrt.be/largest-study-of-its-kind-shows-ai-assistants-misrepresent-news-content-45-of-the-time-regardless-of-language-or-territory) at least one significant issue in 45% of them, with sourcing missing or misleading in 31% and major accuracy problems, including fabricated or outdated details, in 20%. An assistant that invents a date isn't lying, because it has nothing to gain, and that makes it a misinformation machine by definition: Confidently wrong, at scale, with a believer at every link. Readers have noticed, since the Reuters Institute puts trust in AI chatbots' answers at 20% globally, against 37% for news overall. We examined the failure modes in our article on whether you can [trust](https://informed.now/blog/can-you-trust-ai-news-summaries) AI news summaries.

### The honest objection: You can't see intent from the outside

The strongest case against the distinction is that you can't see intent from the outside. On the page, a planted lie and an honest error look identical, and the people best placed to know which is which, the authors, are the last ones who'll tell you. The labels also get used as weapons. Calling a claim "disinformation" accuses its source of lying, so the word gets reached for against inconvenient claims as readily as false ones, and a reader who hears it too often learns to ignore it.

Both points are right, and neither removes the distinction. Intent is established the way the Doppelganger case established it, by investigation after the fact, not by a reader at breakfast. That means the reader's first job is the same for both kinds: Decide whether the claim is true. Intent becomes your question at the second step, when you decide what to do about the claim and how much to trust the next one from the same place. The distinction was never meant to be read off a headline. It tells you what the problem is once you've found one.

### How to sort a false claim yourself

You don't need a fact-checking site or a platform label to run this. It works on a forwarded message, a headline, or a clip, and takes a few minutes.

1. **Find the first source.** Trace the claim back to the document, the dataset, the recording, or the person who said it. A claim with no first source is unverified, whatever its intent.
2. **Name the kind of claim.** A fact can be checked now, an opinion can't be true or false, and a claim about the future can't be checked until the future arrives. Most false claims that spread are predictions or opinions dressed as facts.
3. **Count independent confirmations.** Ten outlets carrying one report is one confirmation, as our article on wire [services](https://informed.now/blog/news-wire-services) explains. Look for a second reporter with a second source.
4. **Ask who gains.** If the claim is false, who benefits from your believing it? Nobody is the usual answer for misinformation. A named someone is the usual answer for disinformation.
5. **Match the response to the answer.** For misinformation, send the correction to the person who shared it, privately, with the source. For disinformation, don't engage, don't quote it to rebut it, and report the account. Either way, stop the chain at you.

The method is the daily version of [news literacy](https://informed.now/blog/news-literacy), and like any habit it holds when you run it on the news you already read.

### Where informed.now fits

Our weekday brief at [informed.now](https://informed.now) is built so that step one is already done: Six news summaries each morning, each citing the strongest source we could reach, which is a primary document in roughly 80% of our coverage. You can follow the link and check the summary against it, which is the only kind of trust worth offering a reader. The method above works without us, on any source you choose.

### Check the claim, then judge the intent

Misinformation and disinformation are the same false claim in two different hands. In the first, the person passing it on believes it, so a better source ends it. In the second, the person knows it's false, so the only thing that ends it is an audience that stops arriving. You can't tell which one you're holding just by reading it, which is why the five checks start with the claim and end with the person. Run them, and neither kind leaves your hands.