How this Indian cybersecurity company created a new AI tool to detect deepfakes

CloudSEK, a prominent cybersecurity firm, has recently uncovered a sophisticated scam involving deepfake videos that exploit the likeness of famous personalities to promote fraudulent mobile gaming apps.

Using cutting-edge artificial intelligence (AI), the company has developed a new deepfake detection tool to identify and combat these scams, which are targeting users across multiple countries.

Deepfake scam explodes
CloudSEK’s investigation revealed that scammers are creating fake videos featuring well-known celebrities endorsing a mobile gaming app called Aviator.

Personalities such as Indian business tycoon Mukesh Ambani, cricket star Virat Kohli, Olympic athlete Neeraj Chopra, and international figures like Cristiano Ronaldo, Ryan Reynolds (as Deadpool), and YouTube sensation Mr. Beast, are falsely portrayed promoting the app.

More from Tech

How ChatGPT is becoming everyone’s BFF and why that’s dangerous America ready for self-driving cars, but it has a legal problem

These videos claim users can make substantial financial gains with minimal investment by playing the game, enticing users into downloading the app. For instance, a deepfake video shows Mr. Beast in a fictitious advertisement aimed at the African market, encouraging users to invest small amounts, like 1,000 Kenyan shillings, to potentially earn large sums through the game.

Virat Kohli has been one of the most frequently targeted celebrities for these deepfake campaigns in South Asia, with videos falsely promoting the app to unsuspecting audiences.

Impact Shorts

More Shorts

America ready for self-driving cars, but it has a legal problem

Alibaba, Baidu begin using own AI chips as China shifts away from US tech amid Nvidia row

Using deepfaked news anchors and journalists
The scammers have also manipulated respected news anchors and channels, to lend credibility to the fraudulent scheme. Fake news broadcasts from trusted platforms like Aaj Tak, Republic TV, Zee News, and ARY News create the illusion of legitimacy, fooling viewers into believing that the mobile app is a genuine opportunity to earn money.

What began as a scam targeting the European Union in early September 2024 has now expanded across multiple regions, including India, Pakistan, Nigeria, Bangladesh, Saudi Arabia, and parts of Southeast Asia.

Interestingly, no deepfake scams were detected in the EU during CloudSEK’s investigation, despite the initial targeting.

Phishing tactics and fake Google Play Store links
One of the most deceptive aspects of the scam is the use of phishing links that mimic the Google Play Store. Users are tricked into believing they are downloading the app from a legitimate source, only to be redirected to fraudulent phishing pages. Domains like “avatarsky[.]one” are used to impersonate the Play Store, and fake apps install malicious software on victims’ devices, displaying real-time statistics and offering payment options through UPI and cryptocurrency to lure users further into the trap.

CloudSEK’s research found that over 1,000 phishing domains are being registered daily, predominantly using the .top domain and originating from Belize. These domains are hosted by an ISP known as IQWeb FZ-LLC, and the scammers offer various payment methods, including bank transfers and cryptocurrency transactions in Bitcoin, Monero, Ethereum, and others.

Deepfake detection tools to the rescue
At the heart of CloudSEK’s efforts to combat this scam is its new free-to-use deepfake detection tool. This tool has been crucial in identifying how cybercriminals are using deepfake technology to exploit high-profile individuals and deceive users worldwide. CloudSEK’s AI-driven tool offers a robust defence against the rise of deepfake scams, empowering users to detect fraudulent content before falling victim.

With cybercriminals continually evolving their methods, CloudSEK’s deepfake detection technology represents a significant step forward in tackling the growing threat of deepfake scams and protecting users from falling prey to these highly sophisticated traps.