Skip to content

In pursuit of hate on the net - Media authority uses AI

Racist comments, defamatory images: There is an institution in Thuringia that combats such content on the net. There are successes, but also frustration.

Radical contents, hate messages and other criminal contents engage the state media institutions as...
Radical contents, hate messages and other criminal contents engage the state media institutions as well (Symbol image)

Hasscrime - In pursuit of hate on the net - Media authority uses AI

In the fight against criminal content on the Internet, the Thuringian Media Authority sometimes successfully uses artificial intelligence (AI), specifically a tool named "Kivi." According to the Media Authority, since May of the previous year, Kivi has facilitated cooperation with sister institutions in other federal states and simplified communication with the Federal and State Criminal Office.

"Kivi supports us so that we no longer have to constantly research the Internet ourselves," summarizes an employee of the Media Authority.

More than 230 problematic contents

The result: 235 deletion requests for hate comments, banned Nazi slogans, defamatory and other criminal contents, whose authors were based in Thuringia, were processed by the Media Authority last year. The majority of these cases involved Facebook. Nearly 90% of these cases were reportedly deleted by the platform after the Media Authority reported and recommended their removal. In addition, the Landeskriminalamt could also be involved in the criminal prosecution of the contents via Kivi.

However, not every platform reacts to deletion notices, according to the Media Authority expert. Effective deletion of reported content was not observed on the Russian network VK or the Telegram Messenger.

"Battle against windmills"

Despite the automated support through AI tools, monitoring content on the Internet for the Media Authority remains a challenge, the employee explains. "Given the flood of content, it sometimes feels like a battle against windmills." Stopping this is not an option.

The AI tool can reportedly examine more than 10,000 pages per day for problematic content. Besides Facebook, other platforms like X, TikTok, and classical websites are also considered.

The Media Authority in North Rhine-Westphalia initiated the Kivi-Tool in 2021 and introduced it, while other Media Anstalten followed suit. According to this, the Community of Media Anstalten with Kivi was able to examine, confirm, and report over 11,600 violations of legal provisions in the period from March 2021 to February 2024 to the Federal Criminal Police Office (BKA).

  1. The fight against criminality on social media platforms like Facebook in Thuringia is aided by the LKA, utilizing the Media Authority's AI tool, Kivi.
  2. The Media Authority in Thuringia processed over 230 problematic contents last year, 90% of which were hate comments, Nazi slogans, and defamatory content, resulting in 235 deletion requests.
  3. Efforts to remove reported content on platforms such as VK and Telegram have not been successful, according to the Media Authority expert, leading to a sense of fighting against windmills.
  4. The AI tool Kivi, initiated by the Media Authority in North Rhine-Westphalia in 2021, has since enabled the Community of Media Anstalten to report over 11,600 violations of legal provisions to the BKA.
  5. Society's increasing reliance on the Internet, including platforms like X, TikTok, and classical websites, has led to an overwhelming flood of content, making the task of monitoring criminality a significant challenge.

Read also:

Comments

Latest