2026 chat room safety report
The Fake Identity Threat: AI, Chat Rooms and Online Safety
Bad actors are using AI to totally fake their identities online. The internet has always had weirdos with confidence, but now some of them have software, fake faces, fake voices and fake conversations. Here is what the current safety data shows, how the scams work, and how to stay safer when you chat.
6,800
AI attack reports in early 2024
440k
Reports by early 2025
23.5k+
Teen blackmail reports by 2025
Global
Online criminals do not care about borders
If a profile looks too perfect, it might be less soulmate and more spreadsheet with cheekbones.
The source
Where does this data come from?
Before diving into the numbers, it is important to know where they come from and what they actually mean. The main statistics in this report come from the National Center for Missing & Exploited Children, also known as NCMEC. It is a large official safety organisation based in the USA.
Even though NCMEC is in America, it receives reports connected to online platforms and tech companies operating around the world. Social media apps, messaging services, chat websites and other online services can send data into that system, so the patterns are not just an American issue.
Criminals online do not care about borders. A scammer in another country can just as easily log into an English chat room and target someone in the UK, Europe or Australia. If a pattern is exploding in official US data, it is sensible to treat it as a warning for everyone using public chat spaces.
AI identity attacks
About 6,800 to 440,000 reports
Reports involving criminals using AI to attack young people rose from about 6,800 in early 2024 to around 440,000 in early 2025. That is not a gentle trend line. That is the internet setting fire to the graph.
Financial sextortion
13,800 to more than 23,500 reports
Reports of teen blackmail linked to sexual images increased sharply between the start of 2024 and 2025, with teenage boys often targeted through fake profiles and private-message pressure.
Global warning
The scam does not stop at borders
The data source is US-based, but the behaviour is worldwide. English-speaking chat rooms, gaming apps and social platforms can all be targeted by people operating from somewhere completely different.
What criminals do with AI
Fake profiles are not just stolen photos anymore.
You have probably heard the phrase generative AI, but in online safety terms it means something very direct: criminals can now create convincing faces, voices, messages and images without needing a real person behind them.
In the past, a scammer or adult creep might steal a random photo from Facebook and use it as a profile picture. That was catfishing. Today, AI can make the fake identity harder to check, because the photo might not exist anywhere else on the internet.
01
Imaginary people
AI image tools can create realistic photos of teenagers who do not exist. Because the person is imaginary, reverse image search often has nothing useful to find.
02
Voice cloning
Cheap software can change how someone sounds in voice chats or voice notes. An adult can make themselves sound younger, which is exactly as grim as it sounds.
03
Chatbots doing the talking
AI chat programs can run hundreds of conversations at the same time, using friendly language and slang until a real criminal decides to step in.
04
Deepfake images
A normal selfie can be turned into a fake explicit image. The original photo might be innocent, but the edited version can still be used as a threat.
The startling truth: because of these new tricks, reported cases of criminals using AI to attack young people rose from about 6,800 in early 2024 to around 440,000 in early 2025. That is a massive explosion of fake identities in chat spaces.
The scam usually starts friendly. So does most terrible advice on the internet.
Teen blackmail
The trap is built to cause panic.
When fake AI profiles, voice tricks and private messaging come together, one of the fastest growing crimes is financial sextortion. This scam targets teenage boys more than anyone else and is growing across the USA and the UK.
The usual pattern is simple and horrible: a criminal uses a fake AI picture of an attractive girl to message a teenage boy in a chat room, gaming app or social platform. They act friendly for a while, build trust, and then push the conversation somewhere private.
1
A fake profile starts the chat
The account looks believable, acts friendly, and may send photos or messages that make the fake identity seem real.
2
A photo becomes leverage
Sometimes the victim sends a real intimate image. Sometimes they only send a normal selfie, and the criminal uses AI to create a fake explicit version.
3
The demand arrives
The scammer demands Bitcoin, gift cards or money, and threatens to send the image to parents, teachers, friends or contacts on Instagram and Snapchat.
This causes extreme panic, which is the whole point of the scam. Reports of this specific blackmail crime jumped from 13,800 to more than 23,500 between the start of 2024 and 2025.
How to stay safer online
You do not need to quit the internet. Just stop giving strangers ammunition.
The numbers sound scary, but you do not need to quit the internet forever. Knowing how criminals operate is the best way to stay safer when using chat rooms, gaming apps or social media.
Never send compromising photos
Even if you really trust the person on the other side of the screen, it is never worth the risk.
Do not trust pictures
A picture or video that matches the story does not mean the person is real. AI can fake convincing material quickly.
Keep it public
Criminals often try to move people away from public rooms and into encrypted private apps where nobody else can see what is happening.
Talk to someone
If you are blackmailed, take a breath and tell a trusted adult or parent immediately. Panic is exactly what the scammer wants.
Do not pay
Paying does not make the problem disappear. It usually proves you can be pressured, and then they ask for more.
You are the first line of defence.
Good chat websites can work on better blocks against bots and AI accounts, but you are always the first line of defence. If someone you just met moves too fast, demands secrets, asks for images, or tries to drag you into a private app, block them and leave. The internet will survive without that conversation. Truly, it has survived worse.
Chat should feel human, not like being hunted by a suspiciously polite robot.
World of Chat is built for people who want public chat rooms with standards, moderation and a bit of common sense. Use the rooms, keep your private information private, and do not let strangers rush you into anything.