…বাংলা
Barlekha Today logo
All the big stories — in your language, updated every 15 minutes
Home › Technology › News
বিজ্ঞাপন
বিজ্ঞাপন — আবদুল কুদ্দুস, চেয়ারম্যান পদপ্রার্থী, ৫ নং দক্ষিণ শাহবাজপুর ইউনিয়ন পরিষদ
Technology

Cloned voices and fake shops: warning on new AI fraud traps and how to stay safe

Staff Reporter • Reading time: 2 min

Criminals are creating new and new fraud techniques using artificial intelligence, or AI. Money is being taken from people by creating fake online shop websites, cloned voices of familiar people and videos that look real, a warning says.

The UK's National Trading Standards has issued fresh guidance to stay alert to such fraud. The body has for the first time published a warning list of the techniques used in fraud committed with AI.

A BBC report said it is becoming possible, using AI, to create professional-looking fake websites in a short time. On those sites, sellers sometimes claim, under the identity of a UK family business, to sell high-quality products made in the country.

Images including UK streets or the country's flag are also used on the websites, so that the shop seems genuine to the buyer. But after buyers order products, they never receive them. Sometimes, instead, low-quality mass-produced goods are sent.

Besides AI-generated fake images, voice-cloning technology and deepfake videos are also being used in fraud. A deepfake is a video created or altered with AI in which someone is seen saying or doing something that did not actually happen.

Another fraud technique is to build a romantic relationship online and take money. In this case, using AI, it is becoming possible to remember a victim's personal information and carry on a conversation for weeks or months.

There are also incidents of money being asked for in an emergency by cloning the voice of a familiar relative or friend. Hearing a familiar voice, many people send money without verifying, and fraudsters take exactly that opportunity.

Using AI, fraudulent emails, text messages and social media posts can be created separately for huge numbers of people. Because those messages contain no spelling or grammatical errors, identifying the fraud is also becoming difficult.

The consumer rights organisation Which? says fraudulent messages can no longer be recognised just by looking for spelling or grammatical errors. Misspelling used to be an easy sign of fraud; AI has almost erased that sign.

National Trading Standards advises that money, gift cards or personal and financial information should not be given to someone who has not been met in person. Whether an online shop's address is real must be checked.

It is also good, the advice says, to confirm whether a job advertisement is genuine by contacting the relevant institution directly. Someone's identity should not be taken as confirmed just by hearing a familiar voice or seeing them on a video call.

If someone pressures you to send money quickly or take an urgent decision, be alert. That is because fraudsters usually try to force the victim to decide without giving them time to think.

Before paying any money, sharing personal information or taking an important step — stop, verify, then decide — is the experts' main advice. In the AI era, those three steps are stressed in the warning as the most reliable way to stay safe from fraud.

বিজ্ঞাপন
বিজ্ঞাপন — জাকির হোসেন, সম্ভাব্য ইউপি সদস্য পদপ্রার্থী, ৭ নং ওয়ার্ড, ৫ নং দক্ষিণ শাহবাজপুর ইউনিয়ন

More in Technology