Deepfake materials are becoming the standard for criminals

Europol has warned of the spread of such content, but fortunately it is still relatively easy to filter.

So-called deepfake materials are computer-generated still and moving images that are made from existing photos and videos. It only takes 500 images or 10 seconds of video to create a completely realistic fake content. Such recordings can be made from virtually anyone who has ever uploaded at least one image of themselves to social networking sites. Journalist Ben Dickson recently wrote about a case in which fraudsters used false faces generated by artificial intelligence to deceive their victims.

Europol’s Innovation Laboratory now brought disclosed in the Facing reality? – Law enforcement and the challenge of deepfakes in which he argued that there was a danger that deepfake content would become a standard tool for criminals. Organized crime in particular is increasingly using computer-generated manipulations. That is why it is very likely that these recordings will play an increasingly prominent role in the commission of simpler crimes in the coming years.

Such substances may also appear in increasing amounts and proportions in disinformation campaigns aimed at influencing or distorting public opinion. They can even portray events in the press that didn’t happen or didn’t happen that way. A good example of this is the disinformation activities used by both sides in the current Russo-Ukrainian war. The authors of the study highlighted that the application of the technique to more serious crimes could gain more and more ground. This includes, for example, phishing campaigns in which perpetrators impersonate company executives and presidents or other senior managers to gain money or important information.

The middle face was created from the two extreme faces, which can be used to create a fake profile, for example

It is fortunate that these contents can still be filtered in real time at their current level of development, for example, blinking or lack thereof can be a telltale sign. Wang Weimin, an employee of Bytedance with the system he developed won the AI ​​Singapore program, led by Singapore State University, in which 470 teams demonstrated their solutions based on artificial intelligence to identify deepfake videos. Wang Weimin stressed that regardless of who thinks about deepfake materials, it is an emerging technology that cannot be ignored. He definitely wants to contribute to the solution of the problem with his solutions based on artificial intelligence.

The specialist he competed alone and the artificial intelligence model he created stood out from the field with 98.5 percent recognition accuracy. His cash prize was S $ 100,000 and he also received S $ 300,000 to create a startup to make a product from his invention. The expert said if his employer is interested in the development, he will integrate it into the BytePlus platform to make it available to users as a service.

Second place went to a team of Swiss software developer Peter Gronquist and Chinese PhD student Ren Yufan, while third came from a group of four Chinese students. The recognition accuracy of the systems they developed also exceeded 98 percent. Singapore Telecommunications Minister Tan Kiat How stated that it is not possible to rely solely on technical solutions to solve technical problems, a wide range of other measures are needed. And one of the key factors in the fight against disinformation is the reliable regional press.

Source: Hírmagazin by

*The article has been translated based on the content of Hírmagazin by If there is any problem regarding the content, copyright, please leave a report below the article. We will try to process as quickly as possible to protect the rights of the author. Thank you very much!

*We just want readers to access information more quickly and easily with other multilingual content, instead of information only available in a certain language.

*We always respect the copyright of the content of the author and always include the original link of the source article.If the author disagrees, just leave the report below the article, the article will be edited or deleted at the request of the author. Thanks very much! Best regards!