disinformation
Intentionally spread false or misleading information aimed at deception, influence, or profit.
In plain terms
Disinformation is intentional deception: someone knows that what they claim is untrue or heavily distorted, yet spreads it because they have an agenda. This agenda can be political influence, money, discrediting a company or person, or simply causing chaos.
It is like someone deliberately turning a road sign in the wrong direction. Those passing by, acting in good faith, follow it and even tell others to go that way. The sign-turner lied, while the others are merely mistaken, yet the damage still occurs.
Artificial intelligence has made this cheaper and faster. Today, plausible fake articles, photos, audio recordings, or videos are created in minutes, even in multiple languages and thousands of variations. At the same time, AI also aids defense: it is used in detecting fake content and coordinated campaigns.
How it works
- Target and audience selection: the distributor decides whom to persuade or sow doubt in, and whose fears, anger or hopes to build on.
- Content creation: a false or distorted story is produced, often today using generative AI-written text, forged images, cloned voices, or deepfake videos.
- Disguising as authentic: the content appears on a news-portal-like site, under the name of someone posing as an expert, or linked to a real event to appear credible.
- Amplification: fake profiles, bots and, increasingly, AI agents that post and comment on their own spread it, while social media recommendation algorithms push content that draws many reactions to even more people.
- Good-faith forwarding: everyday users share content because they believe it. From then on, the lie comes from friends, making it much harder to question.
An everyday example
A video arrives in the family chat group in which a well-known doctor recommends a "miracle cure" and asks you to share it quickly before it gets blocked. The video is an AI-generated forgery; the doctor never said this, and the link leads to an online store. If you forward it, you unwittingly become part of the campaign.
Who is affected
A common misconception
Limits and risks
- AI detection tools are unreliable: they may falsely flag genuine content as fake and let clever forgeries through, so decisions should not be based on them alone.
- Intent is difficult to prove from the outside, so it is often impossible to determine with certainty whether someone is lying or simply mistaken.
- The "disinformation" label is also used as a weapon in debates: not everything is disinformation that someone finds unpleasant or disagrees with.
- If you already consider everything false, that is also a loss: excessive suspicion makes it easy to dismiss genuine evidence with a 'surely AI did it' exclamation.
Similar, but not the same
- misinformation – Misinformation is spread in good faith, without intent to deceive, while disinformation is spread deliberately.
- hallucination – A hallucination is an unintentional error by a language model when it confidently states something false. It becomes disinformation only if someone spreads it deliberately with deceptive intent.
- deepfake – A deepfake is a technique for generating false images, audio, or video, meaning it is a tool of disinformation, not the phenomenon itself.
- propaganda – Propaganda is the one-sided promotion of a view or interest and can use true statements; the essence of disinformation, however, is content that is knowingly false or misleading.
Related concepts
information disorder, information manipulation, AI risks
fake news, disinformation campaign, botnet
misinformation, propaganda, deepfake, voice cloning, hallucination, fact-checking, watermarking, media literacy
A brief history
The word originates from the Russian expression 'dezinformacija' (дезинформация), which entered Western languages from Soviet intelligence terminology during the Cold War. The modern, online interpretation was widely popularized by a 2018 report from the European Commission's expert group. Since the spread of generative AI at the end of 2022, the topic has become one of the central questions in AI debates.
Where things stand
By the end of 2026, AI-generated deceptive text, images, audio, and video will be commonplace and very cheap to produce. The EU aims to contain this through the Digital Services Act (DSA) and its associated Code of Practice on Disinformation, as well as the transparency (labeling) rules of the AI Act applicable from August 2026. However, detection and labeling remain unreliable.
In the news
- Study: a third of young US adults use AI chatbots for information on drugs and alcoholOct 26, 2026 · Education & Family
- Klebelsberg Centre backs Újpetrei primary school targeted by fake news and AI imagesOct 22, 2026 · Education & Family
- Insurers split over AI coverage as CFC launches IP cover and Munich Re's aiSure backs live lending modelSep 23, 2026 · Money
- Content-provenance verification could help against deepfakes and fake newsSep 21, 2026 · Money
- ELTE launches free training for teachers on deepfakes, fake news and AISep 14, 2026 · Education & Family
Also known as
disinformation · disinformation · fake news · fake news · misinformation · disinformation campaign · information manipulation · FIMI
This entry was machine-edited and has undergone machine expert review; the “In the news” section updates automatically from Agent Age stories.