Chapter 4: Harm Online – Harassment, Misinformation, and the Upstander

Harm Online 🛡️

Harassment, Misinformation, and the Upstander

For all its good, the internet can be a harmful place. The same features that let people connect (scale, speed, anonymity, and algorithms that reward strong reactions) also let people hurt others faster and more widely than ever. Cruelty can be crowdsourced. A rumor can circle the world before a correction is written. A single fabricated video can damage a reputation in an afternoon.

This is the hardest chapter in the book, because it deals with how online life goes wrong: harassment, deliberate falsehoods, and synthetic media. It is also the most empowering. In almost every online harm there are more witnesses than attackers, and witnesses can choose to become something better. Understanding how harm spreads, and why people cause it, is the first step in refusing to be part of it and in helping others stop it.

4.1 — The Many Faces of Online Harassment 😠

cyberbullying is the use of digital tools to repeatedly harass, threaten, embarrass, or exclude someone. Because a phone comes home with you, cyberbullying can be relentless in a way schoolyard bullying rarely was: there is no safe place, no end of the school day, and often a large audience watching.

Forms Worth Naming

Recognizing the specific form is the first defense:

Flaming and trolling: posting cruel or inflammatory messages to provoke a reaction.
Harassment: repeated, targeted messages meant to distress someone.
Exclusion: deliberately leaving someone out of a group chat, game, or server to hurt them.
Outing: sharing someone’s secrets, private messages, or personal identity without consent.
doxxing: publishing someone’s private information, such as an address, phone number, or schedule, so others can harass them. It is among the most dangerous forms because it can move harm into the physical world.
Pile-ons: a crowd descending on one person at once, so that each “small” comment adds up to overwhelming abuse.
Sextortion: threatening to share intimate images unless the victim sends money or more images. It often begins with a catfishing account and targets teens of all genders.

Why People Do It: Threat Actors and Motives

A threat actor is anyone who intentionally causes harm in digital spaces. Security professionals analyze threat actors by motive, because motive predicts behavior and the best response. The same approach works for online harm:

Threat Actor	Typical Motive	What It Means for You
Troll	Attention, amusement, a reaction	Not reacting removes the reward; report and move on
Bully or harasser	Power, status in a group, revenge	Document, block, report, and involve adults; support the target
Scammer or sextortionist	Money	Never pay or send more; stop contact, save evidence, report
Disinformation campaign	Political or ideological influence	Verify before sharing; do not amplify
Engagement farmer	Clicks, followers, ad money	Outrage is the product; refusing to share starves it

If It Happens to You

If you are targeted: do not retaliate, because it usually escalates. Save evidence with screenshots that show dates and usernames. Block and report through the platform’s tools. Tell a trusted adult, such as a parent, counselor, or teacher. If someone is threatening to share intimate images of a minor, the U.S. National Center for Missing & Exploited Children offers a free tool called Take It Down that helps remove those images from participating platforms, and law enforcement takes these cases seriously. Reaching out is not weakness. Serious harassment is something no one should handle alone.

The Law Has a Role

Online harassment is not only against platform rules. Every U.S. state has an anti-bullying law, and most explicitly cover electronic or cyberbullying. Credible threats, stalking, and sharing sexual images of minors are crimes, and federal law also prohibits cyberstalking. In 2021, the U.S. Supreme Court ruled in Mahanoy Area School District v. B.L. that schools have less power to punish students for off-campus speech, but it noted schools can still respond to serious bullying, harassment, and threats.

4.2 — Being an Upstander, Not a Bystander 🦸

Most online harm has an audience. When a pile-on happens, far more people watch than participate, and what those watchers do matters enormously.

The Bystander Effect Online

Psychologists have long studied the bystander effect: when many people witness a problem, each one feels less responsible for acting, because they assume someone else will. Online, the effect can be stronger. You cannot see how many others are watching, nobody knows you saw it, and scrolling past costs nothing.

A bystander witnesses harm and does nothing. An upstander witnesses the same thing and acts. You do not need to be a hero or start a fight. Small actions, multiplied, change what a community tolerates.

The Upstander Toolkit

Action	Public or Private	What It Does
Don’t amplify	Private	Refusing to like, share, or comment starves harmful content of reach
Check on the target	Private	A message like “That wasn’t okay. Are you alright?” tells someone they are not alone
Object briefly	Public	“Not cool” from one person breaks a pile-on’s momentum and gives others permission to speak
Redirect	Public	Changing the subject or posting something kind shifts the thread’s energy
Report	Private	Platform reports are how harmful content gets reviewed and removed
Tell an adult	Private	Serious threats, doxxing, or images of minors need adult and sometimes legal help

One caution: do not screenshot and spread someone’s humiliation “to raise awareness.” Even disapproving shares extend its reach. Save evidence to report, not to circulate.

Why It Works

Research on bullying consistently finds that peers are present in most incidents, and that when peers intervene, bullying often stops quickly. Online, the same principle applies: silence reads as approval, and even one dissenting voice changes how others interpret the situation.

Being an upstander also requires judgment about your own safety. You never have to confront someone aggressive directly. Private actions, such as supporting the target, reporting, and telling an adult, are just as valuable as public ones, and sometimes more so.

What to Actually Say

Many people stay silent because they do not know the right words. Short, calm phrases work best, and none of them require insulting anyone:

“Hey, this isn’t funny anymore. Let’s drop it.”
“I don’t think we should be sharing this.”
To the target, privately: “I saw what happened. You didn’t deserve that. Want to talk, or want me to help report it?”
To a friend who joined in: “I know you didn’t mean it badly, but that comment really piled on.”

Notice that each phrase names the behavior, not the person’s character. That makes it easier for others to agree and harder for the conversation to turn on you.

Inclusive Communities Prevent Harm

The best upstanding happens before harm starts. Communities with clear norms, active moderators, and members who welcome newcomers have less harassment to begin with. Including someone who is often left out of a group chat, crediting a quieter teammate’s idea, or setting a friendly tone in a new server are all upstander moves, even though no one is being attacked yet.

4.3 — Misinformation and Disinformation 📰

Not every online harm targets a person. Some targets the truth, and it spreads through ordinary people sharing in good faith. Two terms are essential, and the difference is intent:

misinformation is false information spread without intent to deceive, like a relative who forwards a fake health tip believing it is real.
disinformation is false information created and spread deliberately to deceive or manipulate, such as propaganda, coordinated hoaxes, or scam campaigns.

Disinformation often becomes misinformation as it travels. A campaign invents a false claim, and then sincere people, fooled by it, pass it along.

Why Falsehoods Travel Faster

In 2018, MIT researchers Soroush Vosoughi, Deb Roy, and Sinan Aral published a study in Science examining roughly a decade of news stories shared on Twitter. False stories spread farther, faster, and more broadly than true ones, and the effect was strongest for political news. The researchers found that false stories tended to be more novel and to provoke stronger emotions such as surprise and disgust. Notably, they found humans, not bots, were the main reason falsehoods spread faster.

That finding puts responsibility on you. Every share is a small endorsement. You do not have to fact-check everything you read, but you are responsible for what you amplify.

Who Creates Disinformation, and Why

Disinformation comes from threat actors with clear motives. Governments have run influence operations on social media; the U.S. Senate Intelligence Committee documented, for example, how Russia’s Internet Research Agency used fake American accounts around the 2016 election. Profit-driven sites publish invented headlines because outrage brings ad clicks. Scammers spread fake giveaways and investment schemes. And generative AI has made it cheap to produce fake articles, images, and comments at scale. Media watchdogs have documented a growing number of websites filled largely with AI-generated “news.”

A Practical Defense: SIFT

When a post makes you want to share it instantly, especially if it makes you angry or triumphant, that is exactly the moment to slow down. Digital literacy expert Mike Caulfield’s SIFT method helps:

Stop. Notice the urge to react. Do not share yet.
Investigate the source. Who is behind this? A real news organization, an expert, or an anonymous account created last month?
Find better coverage. Search the claim. Are reliable outlets reporting the same thing?
Trace claims to the original context. Screenshots and clips are often cropped, mistranslated, or years old.

Professional fact-checkers use a related habit called lateral reading: instead of studying one website closely, they open new tabs to see what other sources say about that site and its claim. Researchers at Stanford found that this approach let fact-checkers evaluate sources faster and more accurately than skilled readers who stayed on the page.

Spreading falsehoods has real costs: people make health decisions, vote, and treat their neighbors based on what they believe. When you realize you shared something false, correct it publicly where you shared it. That small act of honesty models the behavior that keeps information spaces healthy.

4.4 — Deepfakes and Synthetic Reality 🎬

The newest and most disorienting form of falsehood is the deepfake: AI-generated or AI-altered media that realistically shows a real person saying or doing something they never did. What once required a film studio can now be done with a phone app, and the results are increasingly hard to distinguish from reality.

Synthetic Media at Scale

Deepfakes matter not just because one fake can fool someone, but because they can now be made in huge numbers by almost anyone. That scale has policy, legal, and societal implications:

Nonconsensual images. Researchers who track deepfakes have found that the large majority of deepfake videos online are sexual content made without consent, overwhelmingly of women. Schools in the United States and other countries have dealt with students using “nudify” apps to create fake images of classmates, causing deep harm to the targets.
Fraud. In 2024, the engineering firm Arup confirmed that an employee in its Hong Kong office was deceived into transferring about 25 million U.S. dollars after a video call in which the other “colleagues” were deepfakes. Voice-cloning scams also imitate family members asking for emergency money.
Elections. In January 2024, a robocall using an AI-generated imitation of President Biden’s voice told New Hampshire voters not to vote in the primary. The Federal Communications Commission then ruled that AI-generated voices in robocalls fall under existing restrictions on artificial voices.
The liar’s dividend. Legal scholars Bobby Chesney and Danielle Citron coined the term liar's dividend for a strange side effect: once people know videos can be faked, wrongdoers can dismiss real evidence as “probably a deepfake.” When nothing can be trusted, truth itself is weakened.

Laws Are Catching Up

Governments have started responding. In May 2025, the United States enacted the TAKE IT DOWN Act, which makes it a federal crime to knowingly publish nonconsensual intimate images, including AI-generated ones, and requires covered platforms to remove them within 48 hours of a valid request. Many states have passed their own laws on sexual deepfakes and on deceptive election deepfakes. The European Union’s AI Act requires that deepfakes be clearly disclosed as artificially generated.

Why Detection Is Not Enough

It is tempting to hope a tool can simply spot fakes. Detection helps, but it has real limits. Detectors trained on one generation of fakes often miss the next, and they can also wrongly flag real media as fake. Compressing, cropping, or re-recording a video can remove the clues detectors depend on. Provenance tools such as Content Credentials, a standard backed by the Coalition for Content Provenance and Authenticity (C2PA), attach a record of how media was made, but that metadata can be stripped when content is screenshotted or re-uploaded. The safest approach combines technology with human habits: verify through trusted sources, and be most skeptical of media that is shocking, urgent, or emotionally perfect.

A Personal Rule

Never create or share sexual or humiliating synthetic images of a real person, and never use someone’s face or voice without consent. If you receive one, do not forward it. Report it and tell a trusted adult. That rule is not just netiquette. In more and more places, it is the law.

4.5 — From Awareness to Action: Building a Safer Community 📣

Knowing about online harm is a start. Changing a community takes planning. This section is about how students, schools, and communities turn awareness into action.

Planning an Awareness Campaign

An effective anti-cyberbullying campaign is more than a poster that says “Be kind.” It follows a plan:

Research the problem. What forms of online harm happen most in your school or community? Anonymous surveys, run with a teacher’s approval, can tell you.
Know your audience. A message for ninth graders differs from one for parents.
Focus on positive norms. Research on the social norms approach suggests people often overestimate how common harmful behavior is. Messages showing that most students do not bully, and do step in, can make upstanding feel normal.
Give specific actions. “Report, don’t repost” and “Check on the target” are more useful than “Stop bullying.”
Make help visible. Share how to report on common platforms and who to talk to at school.
Measure and adjust. Compare survey results before and after, and improve the campaign.

When you present a campaign, lead with a real but anonymous story or a clear finding, show the effects of cyberbullying on targets, bystanders, and even the people doing it, and end with the actions you want your audience to take.

Rules and Agreements That Help

Community agreements matter. Platform community guidelines ban harassment, threats, and nonconsensual images and explain how to report them. School acceptable use policies set rules for school devices and accounts, and many districts have explicit policies on cyberbullying, including off-campus behavior that disrupts school. Knowing these rules lets you use them: a report that cites the specific guideline being broken is more likely to be acted on. Reports also work best with evidence: screenshots that show the username, the date, and the content, saved privately and shared only with the people who can act on them, such as a platform’s safety team, a school administrator, or, for threats and images of minors, the police.

A Framework for Responding to AI-Generated Disinformation

Schools and community organizations increasingly need a plan for AI-generated falsehoods, like a fake video of a principal or a fabricated announcement. A strong framework has five parts:

Stage	What the Institution Does
Prepare	Name a response team, set up official channels people trust, and teach staff and students verification habits
Detect	Encourage quick reporting and monitor for impersonation of the institution and its people
Verify	Confirm with original sources before responding, and avoid repeating the fake in the correction
Respond	Correct quickly through official channels, request platform removal, and involve law enforcement when laws are broken
Support and learn	Care for the people targeted, then review what worked and update the plan

A framework like this has to be defended, not just written. Who decides what is fake? How fast is fast enough? How does the institution protect free expression while removing harmful fakes? Working through those trade-offs is exactly what citizens and leaders must do.

Chapter 4 Activity: The Upstander Campaign 🔍

The Challenge: You will analyze how online harm spreads and design a campaign that helps stop it. Do all research with district-approved tools.

Part 1: The Upstander Playbook

Read this scenario: A classmate posts an awkward video. Within an hour, a group chat is flooded with mocking screenshots, and more people keep joining. Using Section 4.2, write four realistic upstander actions, at least one public and one private, and explain what harm each one reduces.

Part 2: Analyze the Motive

Choose one threat actor from the table in Section 4.1. Describe a realistic (invented, not real-person) situation involving that actor, identify the motive, and explain how knowing the motive changes the best response.

Part 3: Fact-Check a Claim

Find a surprising or emotional claim circulating online. Run it through all four SIFT steps and use lateral reading. Write up what you find and decide whether you would share it.

Part 4: Plan and Present a Campaign

In a small group, plan an anti-cyberbullying awareness campaign for your school using the six steps in Section 4.5. Create one campaign item (a poster, a short slide deck, or a script for a morning announcement) and present it to the class, explaining the effects of cyberbullying you are addressing and the specific prevention actions you are asking for.