19/05/2025

LYFE MONDAY | MAY 19, 2025

22

Cyber threats exploit kids’ brands I N this constantly evolving digital world, cyber threats have taken a different shape by using popular family-focused brands such as Disney, Lego, Toca Boca and others based on o Criminals disguise malicious files as cartoons, games infection attempts were linked to “downloaders”, software that may appear harmless but is often used to silently deliver other potentially dangerous applications.

ensuring that the structure remains physically stable at each stage. The aim is to offer a multipurpose tool, useful in education and design, enabling a project to be designed from a simple text query. Although LegoGPT is still in its infancy, the researchers have made their code and database available, enabling any interested person or entity to explore and use their tool. Details and resources are available on the project’s official website. LegoGPT is part of a wider trend of AI-powered tools for designing Lego constructions. Other projects already well advanced include BrickCenter, a free online platform that uses AI to generate custom Lego models from text descriptions, and Brickit, a mobile application that scans available Lego bricks to suggest feasible constructions. – ETX Studio These downloaders are frequently disguised as games, videos or installers related to popular brands, making them especially effective at tricking users. Furthermore, over 7,800 cases involved “Trojans”, which can steal sensitive data, monitor activity or grant remote access to attackers. These are particularly dangerous when hiding inside seemingly innocent files, such as cheats or fan-made apps. Meanwhile, adware accounted for over 6,400 attempted attacks, typically appearing as flashy games or video apps that bombard users with unwanted ads, slowing down devices and potentially opening the door to additional threats. Fake websites Along with the aforementioned, multiple scam and phishing websites mimicking the design and branding of popular family companies were also identified. One notable example was a phishing page crafted to resemble the official Tokyo Disney Resort website. Such scams are often indistinguishable from legitimate pages at first glance, with the only difference being the URL of the website. The fraudulent site offered users the chance to “buy” park tickets, just like the real one, and prompted them to enter their personal and payment information. However, instead of securing a magical day at the theme park, victims could have their bank card details stolen. Another discovery made involved scams exploiting the name of MrBeast, a popular YouTube celebrity widely followed by children and teens and well-known for giving away expensive prizes such as gadgets, money and even houses. Cybercriminals created phishing pages promising “free gifts from MrBeast”, including digital gift cards for platforms such as Roblox, Xbox and PlayStation. The site prompted users to choose their prize and complete a seemingly harmless task to claim it. To increase urgency, a countdown timer was displayed, urging visitors to “complete a sponsored activity” within a limited time to unlock the final reward code. The entire process is a tactic designed to redirect victims to increasingly deceptive scam pages. Eventually, users are asked to pay a small commission fee to claim their “gift”. However, after submitting the payment, the victim may be

a recent analysis that has revealed a 38% increase in attack attempts from the second quarter of last year to the start of this year. Starting from just 89,000 attempted attacks exploiting these children and family-related brands, it reached almost 123,000 quarter by quarter, with over 432,000 detected in total throughout the period. Lego-themed content accounted for the overwhelming majority of attacks, with over 306,000 attempts, followed by Disney and Toca Boca. Paw Patrol and Peppa Pig were also used as popular lures, though to a lesser extent at 12,500 and 4,900 attempted attacks. Cybercriminals exploit the popularity and emotional familiarity of these brands to trick users into downloading malicious files, often disguised as cartoons or games. The more popular the brand is, the more attractive it becomes as a hook for threat actors. Malicious downloaders The analysis by Kaspersky shows that the most common threats targeting children and families are not always the most obvious ones. Throughout the reported period, nearly 400,000

An example of a phishing page exploiting the brand of Tokyo Disney Resort.

The number of attempted cyber attacks on children from Q2 2024 to Q1 2025.

their device. By staying informed about the latest threats that crop up online and actively monitoring their children’s online activities, parents can create a safer online environment for their kids. With dedicated apps for digital parenting, parents can safeguard their children across online and offline spaces. Such apps help adults ensure a safe and positive digital experience for little ones, establish healthy habits, protect from inappropriate content, balance screen time and even check their children’s physical location.

left with no reward and may have lost money.

What to do To keep children safe online, parents should have an open line of communication with their children about the potential risks they may encounter online and enforce strict guidelines to ensure their safety. Additionally, to keep children from downloading any malicious files during their gaming or regular online experience, parents are advised to install a trusted security solution on

An example of a phishing page exploiting the brand of MrBeast.

This AI can help you build anything with Lego

A team of artificial intelligence (AI) researchers at Carnegie Mellon University, US, recently unveiled LegoGPT, a system capable of generating construction models using the famous little bricks from simple text descriptions. AI helps with the design, while robotic arms build them. LegoGPT is the result of a collaboration between experts in machine learning and robotics. Their aim was to create a tool for transforming text-based instructions into buildable and stable Lego models for hobbyists and professionals alike. The system is based on a language model trained on a vast dataset of over 47,000 Lego structures, combined with highly detailed descriptions. When a query is entered, LegoGPT predicts the addition of each brick in sequence,

Researchers aim to create a tool for transforming text-based instructions into buildable and

stable Lego models for

hobbyists and professionals alike. – PEXELSPIC