What you will learn
- What deepfakes and AI scams look like
- The warning signs
- A simple habit that beats most scams
- Where to get help in Guyana
A voice you know, asking for money
Picture this. It is 9 pm. A voice note comes in on WhatsApp. It sounds exactly like your nephew who works offshore. He is crying. He says he had an accident, his phone is broken, and he needs $80,000 sent by mobile money right now to a "friend's" number.
Ten years ago, a scammer would need to be a good actor. Today, AI tools can copy a person's voice from a few short clips, and can make fake videos of a face speaking. These fakes are called deepfakes.
Guyana's national computer incident response team, CIRT.GY, issued an advisory in September 2025 warning that criminals are using AI to create convincing fake audio, video and images that can pretend to be bosses, co-workers or family members, to trick people into sending money, sharing information or giving system access (CIRT.GY, 29 Sep 2025).
Common AI-assisted scams
- Family emergency voice or video. A cloned voice of a relative asking for urgent money.
- Fake boss. A message, call or video "from the manager" telling a cashier or accounts clerk to pay a new supplier today.
- Fake celebrity or official endorsement. A video that seems to show a well-known person or an official promoting an investment, a "government grant" or a miracle product.
- Perfect phishing messages. AI writes clean, professional emails and texts that look like they come from your bank, GRA, a courier or a shipping company, asking you to "verify" your account.
- Fake online shops and sellers. AI-made product photos and reviews for goods that do not exist.
- Romance and friendship scams. Fake profiles with AI-made photos that build trust over weeks, then ask for money.
- Fake job offers. Well-written offers for jobs abroad or "work from home" that ask for a fee first.
Warning signs
- Urgency. "Right now", "before the bank closes", "don't tell anyone".
- Secrecy. They ask you not to call anyone else to check.
- A new number or new account. "My phone broke, use this number."
- Unusual payment. Gift cards, crypto, or sending to a stranger's mobile money or bank account.
- Something slightly off. In videos: odd blinking, lips not matching words, blurry edges around the face, strange lighting. In voices: flat emotion, odd pauses, background that does not fit. But be careful: good deepfakes may show none of these signs. Do not rely on spotting flaws.
- Too good to be true. Guaranteed profits, free money, prizes you did not enter.
The habit that beats most scams: stop, check, call back
CIRT.GY's advice is to verify unusual requests through a second channel before acting, for example by calling a known phone number or using an official email address (same source). In everyday words:
- Stop. Do not act while you feel rushed or scared. Scammers rely on panic.
- Check. Hang up or stop replying. Look up the person's number that you already had saved, or the official contact on the organisation's own website.
- Call back. Call that known number and ask directly. If it was really your nephew, he will understand.
Family code word
Agree on a simple code word with close family that only you know. Never post it online. If someone calls in an "emergency", ask for the code word.
At work
Businesses should have a clear rule: no payment to a new account, and no change to a supplier's bank details, without a call-back to a known number and a second person's approval. CIRT.GY also advises organisations to train staff to recognise AI-generated content and to set approval steps for sensitive transactions (same source).
Protect your own face and voice
- Think before posting long videos or voice clips publicly. They can be used to train a voice copy.
- Set social media profiles to private where you can.
- Use strong, different passwords and turn on two-step verification for email, banking and social media.
- Never share one-time codes (OTPs) with anyone, even someone who says they are from your bank.
If someone makes a deepfake of you
- Do not panic, and do not pay anyone who threatens to share it.
- Save evidence: screenshots, links, dates, phone numbers.
- Report the content to the platform where it appears.
- Report the matter to the Guyana Police Force.
- Talk to someone you trust. If you need legal options, speak to an attorney-at-law.
What the law covers
Guyana's Cybercrime Act was passed in 2018. The Ministry of Legal Affairs has summarised section 19 as making it an offence to use a computer system to publish or send electronic data that intimidates a person, threatens violence, is obscene, or damages another person's reputation or exposes them to public ridicule, with penalties that can reach $10 million and ten years' imprisonment on conviction on indictment (Ministry of Legal Affairs, 25 Sep 2024). In the same 2024 item, the Attorney General said a review of the Act was underway, and in January 2026 the Prime Minister listed efforts to update the Cybercrime Act among the government's digital priorities (DPI, 29 Jan 2026).
This is a short summary for awareness, not legal advice. Whether a particular act is an offence is for the police, prosecutors and courts to decide.
Do not become part of the problem
- Do not share a shocking video or voice note until you have checked it. Forwarding a fake spreads harm.
- Do not make deepfakes of real people, even "as a joke". It can hurt them and may break the law.
- Do not use AI to write threatening or insulting messages.
Where to get help
- CIRT.GY (Guyana National Computer Incident Response Team) publishes advisories and contact details on its official site: https://cirt.gy/
- Guyana Police Force for crimes and threats.
- Your bank or mobile money provider immediately if you sent money or shared account details. Use the number on your card or their official website.
Key words
- Deepfake: fake audio, video or images made with AI to look and sound like a real person.
- Phishing: a fake message designed to trick you into giving information or money.
- Voice cloning: using AI to copy how a person sounds.
- Two-step verification: a second check, such as a code, when you log in.