Online Harassment Laws by Country: Where Each Route Leads
Online harassment laws by country: how the US, UK, Australia, Canada, Ireland and the EU treat repeated abuse, and which route can get content taken down.
Online harassment laws by country share one core idea and differ on almost everything else. Each system punishes a pattern of unwanted conduct that frightens, alarms or distresses a person, but the pattern required, the harm that must be shown, who can bring the case and whether any route removes the content change from one country to the next.
The practical question is rarely "is this illegal?" It is "which institution can act, and will it touch the posts or only the person?" A criminal charge punishes the harasser. A civil order restrains the harasser. A regulator pressures the platform. Only the last, and sometimes a court order that names the content, reaches the posts themselves.
This article compares the United States, the United Kingdom, Australia, Canada, Ireland and the European Union, using the statutes and regulator pages as they read when we checked them. It stays on harassment and stalking. Defamation, privacy claims and intimate image abuse follow different rules and are outside this comparison.
- Every jurisdiction here criminalizes a course of conduct that harasses or stalks, and online conduct is covered. The tests differ: fear for safety in Canada, alarm or distress in the UK, substantial emotional distress in the US federal statute.
- Criminal and civil routes restrain the person. They do not by themselves delete a post. Removal comes from the platform, a regulator that pressures it, or a court order that names the content.
- Regulators help in a narrow way. Ofcom says it cannot investigate individual complaints. Australia's eSafety Commissioner can act on individual adult cases, but only above a high bar.
- Start with a dated record and a platform report, then choose police or a civil court by country and severity.
Three figures from the primary sources frame the comparison.
Section 7 of the Protection from Harassment Act 1997 (legislation.gov.uk), Ofcom's complaints page, section 264(3) of the Criminal Code of Canada, and section 10 of the Irish Non-Fatal Offences against the Person Act 1997 as revised, all read for this article.
Online harassment laws by country ask the same four questions
Every system asks whether there was a pattern, whether it targeted a person, whether it caused a defined harm, and what the other person intended or should have known.
Pattern. One nasty post is usually not enough. The UK defines a course of conduct as conduct on at least two occasions toward one person. Canada's offence lists acts such as repeatedly communicating, plus monitoring and besetting, which do not use the word "repeatedly." The US federal definition (18 U.S.C. 2266) is a pattern of two or more acts showing a continuity of purpose.
Related readUAE Defamation Law Online: Claim, Complaint, LimitsHarm. This is where systems split most: fear for safety, alarm or distress, or both as alternatives. The level decides whether relentless but non-threatening contact falls inside the law.
Target. Harassment law protects individuals. The UK Act defines "person" in this context as an individual human being, so a company usually has no harassment claim of its own.
State of mind. Most offences ask whether the person intended to harass or ought to have known. Canada adds recklessness. The UK uses an objective test: whether a reasonable person with the same information would think the conduct was harassment.
Criminal, civil and platform routes do different jobs
A criminal route is run by the state. Police investigate, prosecutors decide, and the outcome can be a conviction, a fine, a prison term and sometimes an order restraining contact. You report; you do not control the case. It fits threats, fear for safety, a traceable offender, or conduct that continues despite warnings.
A civil route is run by you, with a lawyer, and aims at a court order and sometimes damages. In England and Wales, section 3 of the Protection from Harassment Act 1997 gives a damages claim and an injunction. You carry cost and the burden of proof, but you control timing.
A platform or regulator route is the only one aimed at the content. It works on the platform's own rules or on legal duties the platform owes over illegal content.
This article is information, not legal advice. Statutes change, courts read them differently, and the facts decide which route fits. A lawyer in the relevant country should confirm any step that carries cost or risk.
Related readGDPR Right to Erasure Exceptions: When Refusal Is LawfulUnited States: one federal cyberstalking statute and many state laws
The United States has a federal provision, 18 U.S.C. 2261A, and a separate harassment or stalking law in each state. There is no national harassment regulator, and platforms are largely shielded from liability for user posts.
Section 2261A covers a person who, with intent to kill, injure, harass, intimidate or place under surveillance, uses the mail, an interactive computer service or an electronic communication service to engage in a course of conduct. The conduct must either place the person in reasonable fear of death or serious bodily injury, or cause, attempt to cause, or be reasonably expected to cause substantial emotional distress. Family members and intimate partners are covered. Under 18 U.S.C. 2261(b), the base-case maximum is a fine and up to five years in prison, higher where injury results. Federal prosecutors enforce it, not the victim.
State laws differ in wording, in how they treat electronic conduct and in severity. As one illustration, New York's aggravated harassment in the second degree (Penal Law 240.30) reaches threats sent by electronic means that the sender knows or should know will cause fear of harm, sent with intent to harass, and is a class A misdemeanor. We did not read other states' provisions, so check the statute of the state involved. Many states also let a civil court issue a protective or restraining order without a conviction, but the name, test and process vary, and the state court's self-help page is the right source.
On content, section 230(c)(1) of 47 U.S.C. 230 provides that a provider or user of an interactive computer service is not treated as the publisher or speaker of information provided by another. Section 230(e)(1) says the section does not impair enforcement of federal criminal statutes. So you or prosecutors can pursue the poster, but a civil claim against the platform over user posts is the weak route. Takedown realistically runs through the platform's harassment policy, backed by a record of the pattern. Whether a court order against the poster can also require removal depends on the state and the judge; we did not read a primary source on that.
Related readDefamation law France: claim, proof and the clockUnited Kingdom: the 1997 Act and the Online Safety Act duties
In England and Wales, the criminal and civil routes both sit in the Protection from Harassment Act 1997, and the Online Safety Act 2023 adds platform duties over illegal content. Scotland and Northern Ireland have their own provisions.
Section 1 bars a course of conduct that amounts to harassment and that the person knows or ought to know amounts to harassment. Section 7 says harassing includes alarming a person or causing distress, and that conduct includes speech. The offences run in tiers:
- Section 2, harassment: summary only, up to six months in prison, a fine up to level 5, or both.
- Section 2A, stalking: harassment involving acts associated with stalking, which include publishing statements or material relating to a person, contacting a person by any means, and monitoring a person's internet or email use. Summary maximum of 51 weeks.
- Sections 4 and 4A: putting people in fear of violence, and stalking causing fear of violence or serious alarm or distress with a substantial adverse effect on day-to-day activities. Each carries up to ten years on indictment.
On the civil side, section 3 allows damages for anxiety and financial loss and an injunction from the High Court or county court. Breach of an injunction is a criminal offence with up to five years on indictment. Section 5A lets a court that acquits a defendant bar conduct where needed to protect someone from harassment. Section 5, the original power for orders on conviction, was repealed on 1 December 2020 by the Sentencing Act 2020. The legislation.gov.uk page does not say what replaced it, so ask a lawyer which order applies.
The Online Safety Act is aimed at platforms. Its Schedule 7 priority offences include sections 2, 2A, 4 and 4A of the 1997 Act, plus Northern Ireland's Articles 4 and 6 and Scotland's section 39 stalking offence. Under section 10, a regulated user-to-user service must run proportionate systems to minimize how long priority illegal content is present and to swiftly take it down once aware of it. Section 21 requires an accessible complaints procedure, including complaints that content is illegal.
Related readGermany defamation online: claims, crime, takedownThe regulator is not a complaints desk. Ofcom states that it cannot respond to or investigate individual complaints, and that reporting directly to the service should be the first step.
Australia: the eSafety Commissioner for adults, and state stalking laws
Australia pairs a federal regulator that can order removal with stalking offences that sit with each state and territory.
The Online Safety Act 2021 (Cth) is the source of the Adult Cyber Abuse Scheme. The Federal Register of Legislation lists section 7 as "Cyber-abuse material targeted at an Australian adult" and section 88 as the removal notice provision. Section 7 asks whether an ordinary reasonable person would conclude the material was likely intended to cause serious harm to a specific Australian adult, and would regard it as menacing, harassing or offensive in all the circumstances. We could not read the operative text of section 88, so for removal we rely on eSafety's own description. It calls the threshold very high.
The platform comes first. eSafety says the quickest way to get abusive material removed is to report it to the service where it was posted. Reports to eSafety are open to people aged 18 or over who live in Australia; under-18s use a separate process. eSafety says its scheme lets it require providers to remove abuse that targets an Australian adult, describes a 24-hour period to comply after a formal notice that it can extend in some cases, and says it can seek civil penalties against a provider that does not comply. One older eSafety page says it has no formal powers here; its scheme pages and the Act's removal notice provision say otherwise. Take exact tests and amounts from the Act.
Related readIndia Defamation Online: Claim, Clock and TakedownCriminal stalking and harassment are mostly state and territory law. The Queensland Criminal Code, for example, has a Chapter 33A in which section 359B defines "unlawful stalking, intimidation, harassment or abuse." We could not read the stalking provisions of Queensland, New South Wales or Victoria in full, so we make no claim about their elements or penalties. For an adult in Australia, eSafety is the only regulator route in this comparison that we confirmed handles individual cases and can direct removal. Its limit is the threshold: ordinary rudeness or persistent criticism does not meet it.
Canada: criminal harassment under section 264
Canada has one national offence and a peace bond procedure. We could not confirm a regulator route for harassing content.
Section 264 of the Criminal Code, headed "Criminal harassment," makes it an offence to engage without lawful authority in listed conduct, with intent to harass or knowing, or reckless as to whether, the person would be harassed, if the conduct could reasonably be expected to cause the other person to believe their safety, or that of anyone known to them, is threatened. Subsection 264(6) states that safety includes psychological safety.
The listed conduct can occur in person, by telecommunication or by other means: repeatedly following; monitoring the person's location, movements, actions or social interactions; repeatedly communicating directly or indirectly; besetting or watching their home or workplace; and threatening conduct. The penalty is up to ten years on indictment, or a summary conviction.
The fear-for-safety element is stricter than "distress." A stream of annoying but non-threatening messages may fall short, and the file has to show why the conduct could reasonably make the person believe their safety is threatened. A parallel route is the peace bond under section 810 of the Criminal Code. Someone who fears on reasonable grounds that another person will cause personal injury to them, their intimate partner or child, or damage their property, can ask a court for a recognizance to keep the peace, lasting up to twelve months and able to carry no-contact conditions. It needs no conviction. It restrains a person; it does not delete a post. Civil claims are largely provincial and are not covered here.
Related readDefamation law South Africa: claim, defenses, clockIreland: an offence and court orders that reach what is said "about" you
Ireland's harassment offence sits in section 10 of the Non-Fatal Offences against the Person Act 1997. The Harassment, Harmful Communications and Related Offences Act 2020 is the Act most often cited and amended section 10, but the Law Reform Commission's revised text shows section 10 was later substituted by the Criminal Justice (Miscellaneous Provisions) Act 2023. Always check the revised text.
On that text, section 10(1) covers a person who, without lawful authority or reasonable excuse, persistently, by his or her acts, intentionally or recklessly seriously interferes with another's peace and privacy or causes alarm, distress or harm, where a reasonable person would realize the acts would have that effect. Section 10(2) is a stalking offence: acts that intentionally or recklessly cause fear of violence, or serious alarm or distress with a substantial adverse impact on day-to-day activities. The listed acts include communicating with or about a person. Summary conviction carries a class A fine, up to twelve months in prison, or both; on indictment, a fine or up to ten years, or both.
Under section 10(4) to (8), the court may order the person not to communicate "by any means with or about" the victim for a specified period and not to approach the victim's home, education or employment. Breach is an offence.
The "about" language is the notable feature. An order can prohibit fresh posts about the victim, which is closer to stopping a campaign than a no-contact order. It still does not tell a platform to remove what is published. That takes a platform report or, for services operating in the EU, the notice-and-action process below.
Related readAnti-SLAPP Law Explained: Early Dismissal and FeesThe EU level: a directive for criminal law, the Digital Services Act for platforms
The EU has no single harassment offence an individual can rely on. It sets floors for national criminal law and regulates how platforms handle reports.
Directive (EU) 2024/1385 on combating violence against women and domestic violence entered into force on 13 June 2024. The European Parliament's legislative summary says it makes cyber stalking, cyber harassment and cyber incitement to hatred or violence criminal offences across the EU, with transposition due by 14 June 2027. A directive binds states, not individuals, so until a country transposes it the national law governs. We make no claim about the elements of its offences.
The Digital Services Act, Regulation (EU) 2022/2065, covers platforms. Per the Commission's questions and answers, platforms must offer an easy way to report illegal content, handle reports promptly, and tell both reporter and poster the decision. A platform that removes or limits content must give a statement of reasons. Users can request review through an internal complaints system, and since 17 February 2024 can use out-of-court dispute settlement. Notices from "trusted flaggers" get priority.
What counts as illegal comes from national or other EU law. The chain for an EU victim is therefore: harassment is a crime under national law, a notice explains why the content is illegal, and the platform must act on the notice. The notice is only as strong as the national offence behind it.
Decision table: which route fits in which country
The table uses the same four criteria for every system. It is a map of routes, not a ranking, and a real file depends on its facts.
Related readCourt order against social media platform, by country| Jurisdiction | Main criminal route | Who can act | Takedown route |
|---|---|---|---|
| United States | 18 U.S.C. 2261A, plus state harassment and stalking laws | Federal or state prosecutors; state civil courts for protective orders | Platform policy; section 230 shields platforms from most civil claims |
| United Kingdom | Protection from Harassment Act 1997, sections 2, 2A, 4, 4A | Police and prosecutors; High Court or county court; Ofcom supervises platforms only | Report to the service under Online Safety Act duties; no individual complaints to Ofcom |
| Australia | State and territory stalking offences | State police and courts; eSafety Commissioner for adults 18 and over | eSafety removal notice if the serious-harm test is met, after a platform report |
| Canada | Criminal Code section 264; peace bond under section 810 | Police and Crown prosecutors; provincial courts for civil claims | Platform report; no regulator route confirmed |
| Ireland | Section 10 of the 1997 Act, as revised | Criminal justice system; court orders on communications with or about the victim | Platform report under the EU Digital Services Act process |
| EU level | Directive (EU) 2024/1385, transposition due 14 June 2027 | National authorities; the EU sets the floor | Digital Services Act notice and action, complaints, dispute settlement |
Three conclusions follow. A takedown never depends on a conviction. Australia's eSafety scheme is the strongest direct tool for an individual adult in this set, while the US offers the least civil leverage over a platform. And in every system, a documented pattern is the one input every route needs.
A realistic sequence once you decide to act
The order matters more than the country. The timings are our experience, stated as usual ranges, and they stretch when a platform appeals or a court has a backlog.
- Preserve the record on day oneSave URLs, screenshots with visible dates and account names, and a log of each occasion. In the UK a course of conduct needs two occasions, so dates matter. Do this before blocking, which can hide content you still need.
- Report to the platform in the first weekUse the harassment form and cite the pattern, not only the latest post. Ofcom and eSafety both ask for this report first, and the EU notice process starts from it.
- Choose police or a civil lawyer within two weeksThreats, fear for safety or a known offender point to police. Relentless but non-threatening contact may point to a civil injunction or protective order where the system has one.
- Escalate a stalled report in weeks two to sixIn Australia, adults can report to eSafety. In the EU, use the platform's internal complaint and out-of-court options. In the UK, use the platform's complaints procedure.
- Expect prosecution and court orders to take monthsA criminal case or injunction hearing usually runs longer than a platform decision. Keep the platform and monitoring work running in parallel.
Two cautions apply throughout. Do not answer the harasser with threats, because it clouds a clear record and can hand the other side a counter-complaint. And do not circulate the harassment widely, because amplification brings the material to more readers and can make a later takedown pointless.
What to do next if this is happening to you now
The next step is the same in every country listed above.
Start the record today and report to the platform, then choose police or a civil lawyer by country and level of fear. If the abuse is already showing when people search your name, a private audit shows what is visible and what is removable. See the private audit.
Where harassment has started to shape what search engines and AI assistants show about you, it is a reputation problem on top of a legal one. Reputation Partners works on that side: documenting what is visible, pursuing removal where a rule or law supports it, and moving accurate, better content above what cannot be removed. The crisis desk covers fast-moving cases, reputation monitoring tracks whether removed posts reappear, and the removal service explains how removal work runs.
Questions people ask
Is cyberbullying a crime in the UK?
There is no single offence named cyberbullying. Repeated conduct that causes alarm or distress can be harassment under the Protection from Harassment Act 1997, and threats or stalking fall under other sections. A one-off unkind post may break platform rules without being a crime.
Does Australia have a law against online harassment of adults?
Yes, in two layers. State and territory stalking and harassment offences deal with the person, and the Online Safety Act 2021 lets the eSafety Commissioner act on cyber abuse aimed at an Australian adult, under a high test of intended serious harm and menacing, harassing or offensive material.
What is the difference between harassment and stalking?
Harassment is the broader idea: a pattern of unwanted conduct that alarms or distresses. Stalking is typically harassment involving fixation behaviors such as following or monitoring. The UK lists examples in section 2A, and Ireland treats stalking as a separate subsection with a higher harm threshold.
Can I sue a platform for harassment posted by a user?
In the United States, section 230(c)(1) generally prevents a provider from being treated as the publisher of another person's content, so civil claims against the platform are hard. In the UK and EU, platforms have duties to act on illegal content, but regulators enforce them rather than individuals in most cases.
Does the law protect a company from online harassment?
Usually not directly. The UK Act defines a person in this context as an individual, and harassment offences turn on fear, alarm or distress felt by a person. The people behind a company may be protected individually, and a company may have other claims that fall outside this article and need a lawyer's advice.
