Deepfake in Zoom and Google Meet, how to Protect your online meetings

21 February 2024 3 minutes Author: Cyber Witcher

Learn how Deepfake technology can infiltrate your Zoom and Google Meet meetings by creating realistic fake videos that can fool participants. Discover protection methods and how to keep your online meetings safe from digital threats.

 

Deepfake technology uses advanced artificial intelligence and machine learning (primarily deep learning) algorithms to create highly realistic video and audio content.

Despite its innovation, Deepfake technology poses significant risks for fraud and scams. By creating realistic video and audio, scammers can impersonate trusted individuals or public figures to manipulate, impersonate or misinform their victims.

There was a recent news story about an “AI scam in which Mayor John Lee purportedly approved an investment plan in a fake video.” Cyber security researchers test software to facilitate diplomacy. If you don’t want to train the models, this software is pretty easy to use. Some of them use existing video files to change faces, others even change faces in real time.

The idea is very simple: you have a camera (we don’t have a webcam for a computer, so we use camera droid, which turns our phone into a webcam). The Deepfake software provides the raw data for the droid camera and has a list of face-swapping models in real-time. We chose John Wick as our POC.

The software can output the merged frame in a new window, which OBS can use to become a virtual camera source.

In Zoom, select the OBS virtual camera as the video source, and then we can use that face to join the Zoom meeting. The video looks smooth, but the audio lags a bit. (this is one of the defects that can be identified from the video of fakes)

As you can see, John Wick has just joined a Zoom meeting.

The researcher  sent several screenshots of the Zoom meeting to various connections, some of whom thought we were meeting real celebrities.

Therefore, we must be ready

Deepke’s scams and scams are becoming more sophisticated, making it harder to distinguish between real and fake content. Here are some tips to help you avoid falling victim to these scams:

  • Be skeptical of video and audio recordings, especially if they contain unusual or sensational content. Analyze inconsistencies in lighting, shadows, or facial expressions.

  • Check the authenticity of the content by checking it with reputable sources. Look for the same news or information on trusted websites.

  • For audio fakes, pay attention to the cadence of the voice, any unnatural pauses, or sounds that seem out of place. This can be an indicator of manipulated content.

  • Be careful about sharing personal information online, especially on social media, as it can be used to create deep fakes or carry out targeted fraud.

  • Some content creators and platforms embed digital watermarks or use blockchain technology to authenticate media. Browser tools and extensions can help you identify these marks.

  • If you come across a deep fake or any content that appears to be fraudulent or fraudulent, please report it to the platform hosting the content and, if necessary, to local authorities.

In a Zoom meeting, John Wick speaks Cantonese fluently

Other related articles
News
Read more
The use of deepfakes in fraud, a case from Hong Kong
The use of deepfakes in fraud is becoming more common, as shown by the incident in Hong Kong, where a financial employee of an international company transferred $25 million to fraudsters who were duped by imitating the appearance and voice of colleagues using deepfake technology
246
Found an error?
If you find an error, take a screenshot and send it to the bot.