# Online Harassment: What to Do, in the Right Order Online harassment: what to do first. Sort the situation, keep a record, cut contact, report to the platform, and know when the law and police apply. --- If you are being harassed online, do not reply. Keep a dated record of everything, cut the person's access to you with blocks and privacy settings, report each item to the platform where it happened, and go to the police if there is a threat, stalking, or any sign that someone may come to you. The order matters, because blocking and reporting make evidence harder to collect if you do them first. Online harassment is repeated, unwanted conduct aimed at one person through messages, comments, tags, fake accounts, posts about them or contact with the people around them. One rude message is unpleasant. A pattern is different: the same person or group keeps coming back, changes accounts, and makes the target feel watched. The pattern is what platforms, police and courts look at, which is why the first job is to make the pattern visible on paper. This guide is written for the person on the receiving end, or for someone helping them. It does not explain how anyone finds or reaches another person, and it uses only invented scenarios. If your address or workplace has been published, the separate rules for being doxxed apply, and a crowd of strangers piling onto one post is a different problem from one persistent person. Rules are described as of the official pages read for this guide, and they change, so check the live version before you file anything. This is information, not legal advice.
The short version
## Online harassment: sort what is happening before you act Start by deciding which of four situations you are in, because each has a different first move. People lose days applying the advice for the wrong one: they ask a platform to act on a threat that belongs with the police, or they call the police about a rude comment that needs a mute button. The table sorts by what the conduct does, not by how upsetting it feels. Being upset is a valid reason to act, but it is not how a platform or a court measures the conduct.
SituationTypical signsFirst moveWho can act
Threat or dangerThreats of harm, mention of your location, travel to you, contact at home or workEmergency services or police, then recordPolice
One persistent personRepeated messages, new accounts after blocks, posts about you, contact with your employer or familyRecord, block, report each item, consider a formal complaintPlatform, regulator, police, court
A group or a crowdMany accounts, a shared post, a hashtag, sudden volumeLimit reach, record the amplifiers, report rule breaksPlatform
Sexual or intimate contentImages shared or threatened, sexualized fakes, sexual messagesUse the intimate-image routes, keep evidence, report to police where requiredPlatform, police, specialist services
If you sit in more than one row, start with the top one. The last row has its own procedures, which this guide does not repeat. It covers what all four share: the record, the cut-off, the report and the escalation. ## Keep a record that a platform, a lawyer and an officer can use A good record has three parts: the content itself, the proof of where it came from, and a log that ties the items into a pattern. Build it before you block, mute or report anything, because blocking can hide a profile from you and a report can lead to content being removed before you saved it. The US government's StopBullying.gov report page gives the core advice for cyberbullying: do not reply to the messages or pass them along, log the date, time and a description of each incident, and save screenshots, emails and text messages, including printed copies, to use when you report.
Build the record in this order
  1. Capture the item as it appearsTake a full-screen screenshot that shows the account name, the text, the date and the platform. Where possible, also save the page or message itself, so the screenshot is not your only copy.
  2. Save the link and the account detailsCopy the web address of the post and the address of the profile. A display name can change overnight; the profile address and the account details in the platform's own export are what let someone trace the account later.
  3. Write one log line the same dayDate, time, platform, what happened, and what you did about it. Keep the log in a plain document that you back up somewhere the harasser cannot reach.
  4. Note the effect, brieflyAdd a line when the conduct changes what you do: a post you deleted, a trip you cancelled, a shift you swapped. These lines show impact, which matters for several legal tests.
  5. Copy it to a second placeSend the folder to a trusted person or a separate, secured account. A single copy on a device that may be compromised is a weak record.
Do not reply to "get it on record": it gives the person a reaction and can expose details about you. Do not crop out the account name or mark up the original. Keep it untouched and annotate a copy. If the harassment includes images or video of you, keep the files securely but do not forward or repost them to "show" anyone. Describe them in the log and let the platform or the police ask for the files. If you are helping someone else, offer to keep the log with them. A second person who can read the messages so the target does not have to see every word is a real help. ## Cut contact without raising the temperature Once the record exists, reduce what the person can reach. Tighten privacy settings, turn off or limit replies and tags where the platform offers it, mute words that are being used against you, and block the accounts you have recorded. Most major platforms offer some version of these tools, and the exact options differ by platform and change often, so look in the settings of each one rather than trusting a list. Blocking is the right default for most people. Its known cost is that a blocked person may escalate, make a new account or switch channel. That is a reason to keep logging after you block, and to treat a new account that behaves the same way as part of the same pattern. Four practical points help: - Change the channel, not just the account. If the person reaches you by phone, email and two platforms, closing one door will not hold. - Check which of your accounts show your real name, employer, school, location or photos, and remove what you do not need to show. - Tell a few trusted people, and ask them not to engage with the person. - Secure the accounts that matter: unique passwords, two-step verification, and a review of logged-in devices. A harasser who can read your messages is more than a harasser. Do not set traps, pretend to be someone else or use anyone else's accounts. Evidence from what is sent to you is sound; evidence from going looking on the other person's side can turn your own record into a problem. ## Report to the platform: what makes a report work Report each item through the platform's own reporting tool, naming the specific rule it breaks. The same site notes that cyberbullying "often violates the terms of service" of platforms, which is why the platform route comes first for most people: it is usually the fastest way to get content down, and it needs no lawyer. Platform policies differ in wording, and the individual policy pages were not all read for this guide, so confirm the current wording on the platform you are using before you rely on any detail. What separates a report that succeeds from one that stalls: - **Name the rule.** Choose the closest category in the form and say in one sentence why the content fits it, such as a repeated unwanted message after a block or content that threatens harm. - **Report the pattern, not only the latest item.** Where a form allows free text, say how many items there are, over what period, and that you have a record. Then report each worst item individually, because a review may look at one item at a time. - **Report the account too.** An account made to harass, or created after you blocked someone, can be reported as a whole. - **Keep your language factual.** A reviewer acts on a clear description of conduct. - **Log the outcome.** Note the date you reported, the reference number if one is given, and what came back. An answer of "no violation found" is still useful: it shows you tried the platform first, which regulators such as eSafety recommend. A refusal is not the end. Many platforms allow an appeal. If the content is on a website rather than a platform, the website owner and the hosting company are the next route. When a platform will not act, a regulator may. Australia has a specific scheme. The eSafety Commissioner's page on adult cyber abuse says it covers content that targets a specific Australian adult and is both intended to cause serious harm and menacing, harassing or offensive in all the circumstances, with realistic threats, repeated targeting of the same person, and sharing personal details or encouraging others to pile on among its examples. It advises saving evidence before you block or delete, recommends reporting to the platform first, since that is usually the fastest route, and says you can report to eSafety if the platform does not help. It says eSafety can direct the person responsible and the service provider to remove seriously harmful content quickly. The page does not state a fixed waiting period or a removal timeframe, so do not assume one. Other countries have different bodies, or none, and the right body depends on where you live. ## When harassment becomes a legal matter Harassment is a legal concept in many places, but the tests differ, and the same conduct can be a crime in one country and only a rule breach in another. Two features recur: the law looks at a course of conduct rather than a single act, and it looks at the effect on the target or the intent of the person. That is why the log in the earlier section is the document that carries the case. Two examples show the shape of the rules, not the whole law of each country; Australia's regulator route is described above. **United Kingdom.** Section 1 of the Protection from Harassment Act 1997, which applies in England and Wales, says in outline that a person must not pursue a course of conduct that amounts to harassment of another and that they know or ought to know amounts to harassment. The test of what they "ought to know" is objective: whether a reasonable person in possession of the same information would think the conduct amounted to harassment. The section has defences, including that the conduct was pursued to prevent or detect crime, was under a legal rule, or was reasonable in the circumstances, and section 1 does not itself define harassment. Other provisions of the Act were not read for this guide, and Scotland and Northern Ireland have their own rules that were not read either, so check the one that applies to you. **United States.** At the federal level, 18 U.S.C. 2261A(2) covers a person who, with intent to kill, injure, harass, intimidate or place under surveillance, uses the mail, an interactive computer service, an electronic communication service or another facility of interstate or foreign commerce to engage in a course of conduct. The course of conduct must either place the target in reasonable fear of death or serious bodily injury, or cause, attempt to cause or reasonably be expected to cause substantial emotional distress to the target, an immediate family member, or a spouse or intimate partner. Federal law is one layer; each state also has its own harassment, stalking and protective order rules, and many cases are handled at state level. Civil routes also exist in many places: a protective or restraining order, an injunction, or a claim for damages. A lawyer can assess them, and can also write a formal letter to a persistent individual. A badly judged letter can provoke the person, so ask for a view on that risk first. If false statements are involved, defamation is a separate route with its own tests, which this guide does not cover. ## Going to the police: what to bring and what to expect Report to the police when the conduct involves threats to harm someone, stalking, hate-motivated conduct, recording you where you would expect privacy, or sexually explicit images or messages, which is the list the US anti-bullying site gives for contacting law enforcement. It also says some states treat other forms of cyberbullying as crimes, so local rules matter. In the UK, Australia and elsewhere the exact offenses differ and were not read for this guide, so check your local rules. If anyone is in immediate danger, call your local emergency number first. Do this before any reporting or documenting. For a non-emergency report, bring a short package rather than a pile: - A one-page summary: who you think is responsible if you know, how it started, how it has changed, and the worst items. - The log, with dates, and the folder of originals. - A list of the platforms and accounts involved, and the dates you reported to each. - The date you asked the person to stop, if you did, and anything that makes you afraid for your safety. Expect questions about whether you know the person and what outcome you want: a warning, formal action or only a record. Ask for a reference number and the next step. Responses vary a great deal by country and officer, and a poor first response does not mean the conduct is not reportable. Go back with new items rather than starting again. If the person is an ex-partner, colleague or neighbor, say so clearly, because that history changes how the conduct is classified and which services can help. ## Protect your search results and your name afterward Harassment leaves traces after the messages stop: posts, threads that name you, fake profiles and screenshots on other sites. Removing a post does not stop a person from posting a new one. For search engines, Google's help page on removing information describes what it will act on under its personal-information policies. These include contact details such as an address, phone number or email, and doxxing, where personal information is shared alongside explicit or implicit threats or calls for others to harm or harass you, or where a significant amount of personal information is gathered without a legitimate purpose. Harassment is not a standalone category on that page. It also says that "even if Google removes something from Google Search, it might still be on the internet", so the page may remain on its host. Full removal means the page does not appear in Search at all, while partial removal hides it only for searches that include your name or another identifier. Other search engines have their own forms, and this guide did not read them. That means a sensible order for lasting traces: 1. Ask the platform or website to remove the content, since that fixes the source. 2. Ask the search engine to stop showing the page if the content matches one of its policies. 3. For fake profiles that use your name or photos, report them as impersonation, which is usually a stronger category than harassment. 4. Keep monitoring your name for a few months, because a persistent person often resurfaces with a new post that quotes the old one. Avoid two moves that usually backfire. A public post "setting the record straight" brings more readers to the thing you want gone. And do not pay anyone who promises to remove harassing content from the web for a fee without explaining the legal basis. In a post dated June 10, 2021, Google described exploitative removal sites as ones that require payment to remove content and said it uses removal requests as a demotion signal against sites with such practices; the lesson is that a fee-for-removal site is a risk, not a shortcut. If your name is attached to a story people search for, a monitoring service can alert you to new results and AI answers. Our reputation monitoring page explains how that works. It shows what appears; it does not make anything stop. ## A realistic timeline and an invented scenario The scenario below is invented. It uses generic parties and assumed facts. It is not a client case and not a prediction. Assume a woman who runs a small online shop is contacted by someone she once turned down for a partnership. Over three weeks the person sends repeated messages, makes two new accounts after being blocked, posts that she cheats customers, and emails her supplier. There is no threat of violence and no mention of her address.
WindowWhat she doesWhat usually happens
Day 1Stops replying, captures messages and posts, starts the log, saves profile linksThe pattern becomes visible on paper for the first time
Days 1 to 2Tightens settings, blocks, reports each item citing the repeated contact after a blockSome content is removed, some is left up; the person usually tries another account
Week 1Logs the new accounts, reports them as evasion, warns the supplier with a short factual noteThe history attached to the report makes the pattern easier to see
Weeks 2 to 3Takes the log to the police as a non-emergency report, asks for a reference numberAn officer may give a warning or open a file; outcomes vary by place and by officer
Month 2 onwardAsks the search engine to stop showing posts that match its policies, monitors her nameContact may thin out if the person has been warned; a lawyer is the next step if it does not
Notice what she does not do: argue in public, contact the person through a back channel, or publish their details in return. Each could weaken a later complaint. The speed in the table is not a promise: the early weeks are usually about reducing reach and building the record, and formal steps take longer. ## Mistakes that keep harassment going Most mistakes come from acting quickly under stress. Knowing them in advance is a protection. - **Replying once "to make it stop".** It rarely does. It shows the person that they get a response, and it adds a reply to a thread you may later need to read out. - **Blocking before capturing.** If you lose sight of the content, you lose the record. - **Deleting your own posts in a panic.** Remove them if you wish, but save them first, because the person may quote or alter them. - **Going public with the person's name.** Naming someone, even correctly, can expose you to a claim and gives the person a reason to escalate. If a name has to be given, it goes to the police or a lawyer. - **Asking a crowd to help.** Friends who reply or contact the harasser can start a second dispute. Give them tasks: read messages for you, keep the log, check on you. - **Treating silence as the end.** A quiet period is not proof the matter is over. Keep the log open and your settings in place for several months. - **Facing it alone.** The effect on health, sleep and work is real. Tell someone close, and use professional support if it wears you down. ## Your next step: one record, one cut-off, one report Today, do three things: start the log, tighten the settings that expose you, and file one report for the worst item through the platform's own tool. If there is a threat, call the police first.
If this is happening to you now

If the harassment is spreading, tied to your name in search results, or aimed at your work, our crisis desk can help you sort what to report, what to record and what to leave alone. Reputation Partners works on search and AI answers around your name, not on the police or legal steps, which stay with the right authorities and your own counsel. Start with the crisis management page.

## Questions people ask ### Is online harassment a crime? It can be. Many countries and states have offenses covering harassment, stalking or threats, and some name online conduct specifically. Whether your case qualifies depends on the place, the pattern, the intent and the effect on you. A police report or a lawyer's view will tell you which rules apply. A platform can act on a rule breach even when the conduct is not a crime. ### Should I block the person or leave them unblocked to collect evidence? Capture first, then block. You can collect evidence from what has already been sent to you and keep logging anything new that gets through from other accounts. Staying exposed to the messages only to gather more usually costs you more than it gives. If you need to keep watching, ask a trusted person to read the account for you. ### What if the platform does not remove it? Appeal if the platform offers a review, and attach the log. If the content breaks the law, report it to the police or a regulator, where your country has one. If it is a post on a website, contact the site owner and the hosting company. Keep a note of each attempt, because refusals still show you tried the fastest route first. ### Can I find out who is behind an anonymous account? Do not try to unmask someone yourself. A wrong guess can harm an innocent person, and amateur tracking can break the law or platform rules. Give the account details and your log to the police or a lawyer; in some places a court can require a platform to disclose account information, and the rules vary by country. ### Will harassment show up when someone searches my name? It can, if the posts name you and appear on a page that search engines index. Ask the website to remove the post and ask the search engine to stop showing it where its policies cover the content. Search removal does not delete the page from the web. Monitor your name for a few months, since persistent people often repost.