Staying Safe in the Age of AI: Fake Schemes, Deepfake Fraud and Privacy

AI has made many tasks easier, but it has also made life easier for fraudsters. Copying someone's voice, producing a fake video and writing messages that look completely genuine are now far easier than ever before. In this article you will learn how to recognise fake "AI earning" schemes, fake apps, voice-clone and deepfake fraud, how to secure your accounts, what you should never type into a chatbot, and how to verify a news story or video. Be sure to share these points with the elders and children in your family too.

1. Fake "AI earning" schemes

You often see ads on social media like: "Earn daily with an AI trading bot", "Just like videos and get paid", "Invest in an AI app and double your money". Most of these are frauds.

Warning signs:

• A promise of certain or fixed profit. No legitimate investment gives a guarantee.
• "Pay a registration fee first" or "Deposit money to activate your account".
• Commission for bringing in new members, meaning the money comes from new people, not from a product.
• At first you get a small amount back to build your trust, then a large amount gets stuck.
• Pressure to hurry: "Today is the last day".
• Fake videos of celebrities shown praising a scheme. These are often deepfakes.

The legitimate way to earn with AI is to learn a skill and do real work, as explained in our article Freelancing with AI. Money is not made by pressing a button.

2. Fake apps and websites

Many fake apps and websites are created using the names of ChatGPT, Gemini or other popular tools, and they steal money or data.

• Always install apps from the Google Play Store or Apple App Store; never download an APK file from a WhatsApp link or a website.
• Even on the Play Store, check the developer's name. The official ChatGPT app is by OpenAI and the Gemini app is by Google LLC.
• Read the website address carefully. Look-alike names, such as an extra letter or a strange domain, are signs of a scam.
• "Cracked" or "free Premium" apps often contain viruses.
• When installing, check which permissions the app is asking for. Why would a chatbot app need your SMS or contacts?

3. Voice-clone and deepfake fraud

Today a person's voice can be copied from a recording just a few seconds long. Fraudsters call or send a voice note in a relative's voice: "I'm in trouble, send money right away, don't tell anyone". In the same way, someone's face can also be faked in a video call or a video.

How to protect yourself:

1. Stop and verify: Hang up and call that person yourself on their own known number.
2. Family code word: Agree on a secret word with your family that can be asked in an emergency. Never write this word anywhere online.
3. Ask a personal question: Something only that person would know.
4. Pressure to hurry and keep it secret is itself a warning bell.
5. Signs in a video: Lips not matching the voice, strange blinking, blurry edges around the face. But remember that a good deepfake may show none of these signs, so verification is the real solution.

Try not to post too many videos and voice recordings of your children and family publicly.

4. Protecting your accounts

Turn on 2-Step Verification

This is the most important step. Along with your password, it requires a second proof, such as a code or prompt sent to your phone.

• Google account: Open myaccount.google.com, go to Security and turn on 2-Step Verification.
• WhatsApp: Go to Settings → Account → Two-step verification and set a six-digit PIN.
• Facebook and Instagram: In Accounts Center, under Password and security, turn on two-factor authentication.

Menu names can change with the app version, so if you have trouble, search for "two-step verification" in the app's Help Center.

Never share your OTP

An OTP (One-Time Password) is the code that arrives by SMS. No bank, mobile company, WhatsApp or government office will ever ask you for an OTP over the phone. Anyone who asks for it is a fraudster, even if they claim to be a bank officer, the police or someone from a "prize scheme". Giving anyone your six-digit WhatsApp code can hand your entire account over to them.

Strong passwords

• Use a different password for every account.
• Make your password long, for example a combination of a few unrelated words.
• Use a trusted password manager such as Google Password Manager.
• Keep your recovery email and phone number up to date.

5. What not to type into a chatbot

Whatever you type into ChatGPT, Gemini or any AI chatbot is stored on the company's servers, and in some cases it may also be used to improve the model. So don't type these things:

• Passwords, OTPs, PINs
• CNIC number, passport number, bank account or card details
• Medical reports or other people's personal information
• Confidential office or client data, unless your organisation has allowed it

If you need help with a document, remove names, numbers and addresses first. Go into each tool's Settings and check what options it offers for chat history and for using your data for training. More guidance for students is in ChatGPT and Gemini for Students.

6. How to verify information

AI itself can write wrong things, and fake AI-made news and images also circulate on social media.

1. Find the original source: Is the story on a reliable news organisation's site or an official government website?
2. Verify images: Search the image with Google Lens to see where and when it was published before.
3. Check the date: Old videos are often spread as if they were new news.
4. Emotional pressure: If a post makes you very angry or afraid, think twice before forwarding it.
5. AI answers: Always check important information, such as health, law and money, with an expert or an official source. To ask better questions, also see how to ask AI for its sources in Prompt Engineering.

What to do if you are scammed

• Immediately call your bank or wallet helpline and get your account or card blocked.
• Change your passwords and log out everywhere.
• File a complaint with the National Cyber Crime Investigation Agency (NCCIA). Its official website, nccia.gov.pk, has the online complaint process and helpline number; these details can change, so always confirm them on the official website.
• Keep screenshots, numbers, transaction IDs and messages safe.
• Warn your family and friends, especially if your account has been hacked.

Summary

• Guaranteed earnings, upfront fees and pressure to hurry are the biggest signs of fraud.
• Install apps only from the official store, and check the developer's name.
• Before believing a voice or video asking for money, call the person yourself to verify, and keep a family code word.
• Turn on 2-Step Verification on every account and never give your OTP to anyone.
• Don't type personal or confidential information into chatbots, and check every important news story with its original source.