Misinformation is false or misleading information that is shared without an intent to deceive: the person spreading it usually believes it is true or has not checked it. In online safety law and wider tech regulation, the concept matters because most legal systems protect the expression of mistaken beliefs, so regulators target the systems that amplify false content rather than the individuals who repeat it.
The term sits next to disinformation, which is deliberately fabricated or manipulated content, and malinformation, which is genuine information used to cause harm. The boundary between these categories is defined by intent and context, not by the content alone, and that single fact shapes how platforms, courts and regulators respond.
How misinformation differs from related information harms
Misinformation is distinguished from its neighbors by the state of mind of the sharer and by whether the content is false at all. The widely used framework of information disorder separates three categories that often blend in practice, because a fabricated claim created as disinformation becomes misinformation once ordinary users share it in good faith.
| Category | Is the content false? | Intent to harm or deceive | Typical example |
|---|---|---|---|
| Misinformation | Yes, or misleading | No | A relative forwards a fake health cure |
| Disinformation | Yes, or manipulated | Yes | A coordinated campaign spreads invented election fraud claims |
| Malinformation | No, it is genuine | Yes | Leaked private messages published to humiliate someone |
| Satire or parody | False on its face | No, if labeled | A humor site publishes an obviously absurd headline |
Because intent cannot be read from a post, platforms rarely classify individual items with certainty. Content moderation teams therefore assess observable signals instead: the potential for real-world harm, the reach of the content, coordinated behavior and whether independent fact-checkers have rated the claim.
Why misinformation spreads on online platforms
Misinformation spreads because engagement-optimized recommender systems reward content that provokes strong emotion, and false claims are often more novel and surprising than accurate ones. Research on social networks has repeatedly found that novelty and outrage predict sharing better than accuracy does.
- Recommender amplification: ranking algorithms push high-engagement posts to users who never followed the original source.
- Low-friction sharing: one-tap resharing removes the pause in which a person might check a claim.
- Context collapse: screenshots and clips travel without their original date, source or caveats.
- Closed messaging groups: end-to-end encrypted channels limit visibility for moderators and fact-checkers.
- Generative AI: synthetic images, audio and text lower the cost of producing plausible false material.
The harm profile varies with the topic. Misinformation about public health, elections, emergencies and financial products causes measurable damage, while trivial errors about celebrities rarely justify intervention. Regulators therefore focus on systemic risk categories rather than on falsity in general.
How regulation treats misinformation
Most democratic legal frameworks do not prohibit misinformation as such, because false but lawful speech is protected by free expression guarantees such as Article 10 of the European Convention on Human Rights and the First Amendment in the United States. Instead, laws impose procedural duties on platforms to assess and mitigate the risks that their services create.
The following table compares how major regimes approach the issue; the details change as implementing acts and codes are updated, so the overview describes the general regulatory model rather than every obligation.
| Framework | Jurisdiction | Approach to misinformation | Who is mainly obliged |
|---|---|---|---|
| Digital Services Act | European Union | Systemic risk assessment and mitigation, including risks to civic discourse and public health | Very large online platforms and search engines |
| Code of Conduct on Disinformation | European Union | Voluntary commitments integrated into the DSA framework, covering demonetization and fact-checking | Signatory platforms and advertisers |
| Online Safety Act | United Kingdom | False communications offense for knowingly false harmful messages; Ofcom advisory committee on misinformation | Individuals and regulated services |
| Section 230 | United States | Platforms are largely shielded from liability and moderate under their own policies | No general statutory duty |
A recurring design principle is that the knowing element is reserved for criminal law. The UK false communications offense, for example, requires that the sender knew the message was false, which places genuine misinformation outside its scope. The duty of care model instead asks services to design systems that reduce harm at scale, which also tends to be documented in each transparency report.
Countermeasures platforms use against misinformation
Platforms address misinformation with a graduated set of interventions that range from adding context to removing content. Removal is usually reserved for content that creates a serious risk of imminent harm, while softer measures apply to disputed but lawful claims.
| Intervention | How it works | Main benefit | Main limitation |
|---|---|---|---|
| Fact-check labels | Independent reviewers rate a claim and the platform attaches the rating | Preserves speech while adding context | Slow and limited to a small share of posts |
| Community notes | Users add context that is shown when raters with differing views agree | Scales and gains perceived neutrality | Notes often appear after peak spread |
| Downranking | Reduces distribution in feeds and recommendations | Limits reach without deletion | Opaque to users and hard to audit |
| Prebunking | Short messages teach manipulation techniques before exposure | Builds resilience across topics | Effects fade without reinforcement |
| Removal | Content is taken down under platform policy or legal order | Stops further spread | High risk of overreach and censorship claims |
Evidence suggests that friction prompts, such as asking users whether they want to read an article before sharing it, measurably reduce resharing of unverified links. No single measure is sufficient, and effective programs combine detection, context, design changes and media literacy.
Limitations and risks of regulating misinformation
The central risk in regulating misinformation is that truth is sometimes contested or still emerging. Claims labeled false early in a crisis have later been reassessed, and rules that compel removal can entrench official positions and chill legitimate debate.
- Definitional vagueness: laws that ban “false news” without precise criteria have been used against journalists in several countries.
- Over-removal: heavy penalties encourage platforms to delete borderline content to avoid liability.
- Measurement problems: prevalence estimates vary widely depending on sampling and definitions.
- Uneven enforcement: moderation capacity is weaker for smaller languages and regional markets.
- Backfire effects: visible removal can lend credibility to a claim among audiences who distrust institutions.
For these reasons, current best practice favors transparency, auditability and user empowerment over direct speech restrictions, with legal takedown obligations kept for clearly illegal content.
The simple explanation
Misinformation is wrong information shared by people who think it is right. Think of a neighbor who passes on a rumor about a road closure without checking it: the neighbor is not lying, but the message is still false and can still cause trouble.
Online, such rumors travel far faster because apps reward posts that excite people. Laws in most democracies do not punish people for being mistaken. Instead, they ask large platforms to notice where false claims could hurt people and to slow their spread with labels, context and better design.
How it is used in practice
During a regional flood, a photo from a different country, taken years earlier, circulates with a caption claiming that a local dam has collapsed. Thousands of residents share it in good faith, and emergency phone lines are overwhelmed within an hour.
A very large platform operating in the European Union applies its crisis response protocol under its Digital Services Act risk mitigation plan. Its systems match the image to the older original, attach a fact-check label with the correct date and location, and reduce recommendation reach for reshared copies while leaving the posts visible. Local authorities publish an official update that the platform pins in search results for the region.
Resharing of the photo drops sharply after the label appears. The platform later documents the incident, the measures taken and their effect in its systemic risk report, which independent auditors review the following year.