Black October, explainedAppsGuidesAI DictionarySmall FixesOn RecordAboutFollow the lab

Misinformation

Also called disinformation, fake news, AI slop

Misinformation is false information shared as if it were true. AI makes it cheaper and faster to produce, and harder to tell from the real thing.

Safety and societyEveryday

In plain words

False stories are as old as gossip. Researchers split them into two kinds. Misinformation is false information shared by people who believe it, like an uncle forwarding a fake cure to the family group. Disinformation is false information spread on purpose to mislead.

AI changes the scale, not the basic problem. A chatbot can write a hundred fake news articles in a minute. Image and voice tools can make a convincing deepfake of a real person. Floods of cheap, low-quality AI content are now called "AI slop".

AI can also spread mistakes by accident. Chatbots that search the web can repeat false claims they find, and can make up facts of their own, called hallucinations. The best defence is still old-fashioned: check where something came from before you believe or share it.

Everyday example

The forward that came too fast

A photo lands in your building's WhatsApp group: a flooded flyover in your city, "happening now". It looks real. But there is no news report, the street signs are blurry, and the cars look oddly smooth. A quick reverse image search finds nothing older than an hour. Someone made it with AI. Twelve people have already forwarded it, each believing it. That is misinformation, started by disinformation.

Where the example breaks: In this example the fake had tell-tale flaws. The best AI images and voices now often have none you can spot, so checking the source matters more than inspecting the picture.

Go deeper

How it actually works

AI affects misinformation in four main ways.

  1. Cheaper, faster, more convincing

    Generative AI lets one person produce text, images, audio and video in volume, in many languages, at almost no cost. Fake sites filled with AI-written articles, sometimes called content farms, can look like real news outlets.

  2. Chatbots that repeat false claims

    In an August 2025 audit by NewsGuard, ten leading chatbots repeated false claims about news topics 35% of the time, up from 18% a year before. As they started searching the web live, they stopped declining to answer and pulled in unreliable sources.

  3. The liar's dividend

    When fakes are everywhere, real evidence becomes easier to deny. Legal scholars Robert Chesney and Danielle Citron named this the liar's dividend in 2019: someone caught on genuine video can simply claim it is a deepfake.

  4. Why false news spreads

    A 2018 MIT study of Twitter from 2006 to 2017 found false stories were 70% more likely to be retweeted than true ones, and people, not bots, drove most of the difference. AI did not create that, but it adds more fuel.

  5. Defences

    Labels and watermarks on AI content, AI detectors (which are far from perfect), fact-checkers, platform rules and laws such as India's 2026 rules requiring labels on AI-generated content. For individuals: pause, check the source, search for other reports, and do not forward what you cannot verify.

Definitions
Misinformation: false, shared without intent to harm. Disinformation: false, shared knowingly to cause harm (Council of Europe report, 2017)
Global risk ranking
Misinformation and disinformation ranked the top global risk for the next two years in the WEF Global Risks Report 2024, and second in 2026. In 2024 it ranked as India's top risk too
Word of the year
Merriam-Webster chose "slop" for 2025: "digital content of low quality that is produced usually in quantity by means of artificial intelligence"
Elections in 2024
Meta said AI content was under 1% of fact-checked election misinformation on its apps. Alan Turing Institute researchers found no conclusive evidence AI swung any major vote

Has AI made misinformation much worse?

One view

Yes, it is a new kind of threat

Fakes are now cheap, endless and hard to spot, and chatbots repeat false claims at scale. Experts surveyed by the World Economic Forum rank misinformation among the biggest short-term global risks, and the liar's dividend lets people dismiss real evidence.

The other view

It is mostly an old problem

False stories spread fast long before AI, driven by people, not machines. In 2024's many elections, AI content was a small share of fact-checked falsehoods, and researchers found no conclusive evidence it changed any result.

Most likely: Both are partly right. So far, AI does not appear to have swung elections, but it raises the volume of junk and makes trust harder. The biggest effect may be people doubting everything, including what is true.

Questions people ask

What is the difference between misinformation and disinformation?

Intent. Misinformation is false information shared by someone who believes it is true. Disinformation is false information created or spread on purpose to deceive or harm. The same fake can start as disinformation and become misinformation as ordinary people forward it.

What is AI slop?

A name for the flood of low-quality content made cheaply and in bulk with AI: bland articles, odd images, recycled videos. Merriam-Webster made "slop" its 2025 word of the year. Slop is not always false, but its sheer volume makes good information harder to find.

How can I tell if a news story or image is made by AI?

Often you cannot tell by looking. Check the source instead: is it from a known outlet? Do other trusted outlets report it? A reverse image search can show if a picture is older or new. Look for AI labels or content credentials. When unsure, do not forward.

Can chatbots give me false information?

Yes. They can make things up (hallucinate) and repeat false claims found online. In a 2025 audit, leading chatbots repeated false news claims about a third of the time. Check anything important, especially health, money and news, against a trusted source.

Did AI change the result of elections?

There is no conclusive evidence that it did in 2024, according to researchers at the Alan Turing Institute, and Meta said AI made up under 1% of fact-checked election misinformation on its apps. Researchers still warn that AI erodes trust in what voters see.

Further reading

Prepared by Mundhra Labs. Reviewed by Aditya Mundhra. Checked October 2026.