Cybersecurity News that Matters

Cybersecurity News that Matters

South Korean police to fight deepfakes ahead of general election

Designed by Areum Hwang, The Readable

by Kuksung Nam

Mar. 05, 2024
11:20 AM GMT+9

South Korean authorities are ramping up efforts to combat deepfake videos, recognizing these deceptive materials as a significant threat to the integrity of the country’s general election scheduled for April.

The Korean National Police Agency (KNPA) unveiled on Tuesday its adoption of new deepfake detection software, which will be immediately integrated into their investigative processes. The agency clarified that the time required to determine whether digital content has been artificially generated using AI technology could range from 5 to 10 minutes, depending on the video’s length and quality.

The newly developed software is based on a dataset comprising 5.2 million data points from 5,400 individuals, including 1 million data points related to South Korean nationals. This initiative aims to address the limitations of current detection tools, which often struggle with accuracy concerning South Korean subjects due to a lack of sufficient data. While the detection tool is intended to serve as a guide during investigations, it has demonstrated an 80% accuracy rate. This implies that for every 10 cases flagged as fake, two might actually be authentic.

Since last June, the police have collaborated with one of the country’s leading artificial intelligence companies to create software aimed at aiding law enforcement in identifying sexually explicit content that has been digitally altered using AI. Given the rapid advancements in technology, which now allow the creation of highly convincing deepfakes of politicians indistinguishable to the naked eye, the South Korean police have decided to expand the application of this new tool. This decision aims to address and mitigate potential risks to the integrity of the upcoming general election.

With the implementation of the revised election law last January, which bans the use of deepfake technology in political campaigns starting 90 days before the election date, the police are set to intensify their efforts in investigating election-related disinformation with the assistance of the new software. An official from the National Office of Investigation at the KNPA remarked, “This is the first deepfake detection software developed by the police. Even last year, deepfake videos were relatively crude, making them easy to identify. However, technology is advancing rapidly, necessitating the development of tools to detect such content.”

As South Korea approaches its general election, the government is increasingly sounding the alarm over election disinformation that could potentially influence voter decisions. According to an exclusive report by the local broadcaster SBS on March 4, the South Korean presidential office engaged in discussions with its United States counterpart last February. The talks focused on collaboration to combat fake news, highlighting the international concern over the impact of disinformation on democratic processes.

South Korea has been collaborating with the U.S. in the battle against disinformation. In December of the previous year, the two nations signed a memorandum of understanding. This agreement stated that both parties would work together to counteract the spread of false information and information operations orchestrated by foreign countries, underscoring their commitment to safeguarding the integrity of information and bolstering defenses against malicious external influences.

Readable Subscription Form - South Korean police to fight deepfakes ahead of general election

Subscription

Subscribe to our newsletter for the latest insights and trends. Tailor your subscription to fit your interests:

By subscribing, you agree to our Privacy Policy. We respect your privacy and are committed to protecting your personal data. Your email address will only be used to send you the information you have requested, and you can unsubscribe at any time through the link provided in our emails.

  • Kuksung Nam
    : Author

    Kuksung Nam is a journalist for The Readable. She has extensively traversed the globe to cover the latest stories on the cyber threat landscape and has been producing in-depth stories on security and...

  • Arthur Gregory Willers

    Arthur Gregory Willers is a copyeditor at The Readable, where he works to make complex cybersecurity news accessible and engaging for readers. With over 20 years in education and publishing, his exper...

  • Areum Hwang
Stay Ahead with The Readable's Cybersecurity Insights