Far-Right Extremists Harness AI-Generated Music to Disseminate Hate on Streaming Platforms
In an unsettling development, far-right extremists are employing AI-generated earworms to propagate their divisive rhetoric across various platforms, including Spotify and TikTok. This burgeoning phenomenon has now secured a prominent position on the Dutch Spotify charts.
The narrative surrounding the rise of AI-generated far-right extremism transcends mere American borders. In fact, it extends into the heart of Europe, particularly in the Netherlands and France, where such music is proliferating alarmingly on digital music platforms.
Recent reports indicate a surge in AI-generated far-right content throughout Europe—evident during last year’s French elections and persisting thereafter. According to Reddit users, there is a “disturbing trend” of AI-generated songs disseminating anti-immigrant sentiment, consistently rising in the Dutch Spotify rankings.
“At present, one of these AI anti-immigrant tracks is nestled within the Dutch Top 5, with approximately eight out of ten of the most viral songs falling under this same vile category,” remarked a Reddit user known as EuroMEK.
“These are not creations of genuine artists; they are AI-generated compositions brimming with extremist themes.”
“What troubles me greatly is Spotify’s tacit approval of this content, despite its virulence,” the user continued.
“How can they rationalize the platforming of AI-generated hate speech that competes with authentic artists on their charts?”
New Insights on Content Longevity from Recent Research
Recent findings from Cornell University revealed a troubling prevalence of extremist audio within “thousands of clips from German, British, and Dutch TikTok feeds.” Over three-quarters of videos featuring such content remained accessible four months post-upload.
This trend’s insidious nature lies in TikTok’s “use-this-sound” feature, which serves as a covert vessel for disseminating hate speech.
Marcus Bösch from Heinrich-Heine University in Düsseldorf highlighted the existence of numerous trends where innocuous memes, such as song-guessing games, mask “brutal, racist, misogynist, and death-themed lyrics” embedded in the music utilized in these trends.
Not all of these songs are AI-generated, yet many associate “hateful messages” with iconic tracks from the ‘90s, such as those by Aqua or Gigi D’Agostino.
“There’s Nazi techno, Nazi pop, Nazi folk—there’s something for everyone,” Bösch noted, emphasizing that the intent behind this tactic is to redirect users to off-platform content designed to indoctrinate them into extremist ideologies.
Moderation on TikTok appears inadequate for tackling audio-based hate content, whereas text-based hate speech is often removed swiftly. Even overtly offensive material—citing instances of a Hitler speech appearing in over 1,000 videos—can elude detection for extended periods.
“It is undeniable that such content is easy to see, hear, or perceive,” he contended. “Locating it should not pose an insurmountable challenge.”
A TikTok spokesperson previously indicated that the platform employs a mix of technology and human oversight to identify and eradicate content promoting hate ideologies, asserting that 94% of such problematic material is removed prior to reporting.
As artificial intelligence evolves, those who misuse it to disseminate hateful narratives will likely become increasingly adept at circumventing detection mechanisms.
Nonetheless, the question remains: can legislation addressing unethical AI effectively mitigate the proliferation of AI-generated hate speech?
Regrettably, the answer seems bleak. While legal frameworks desperately require modernization to keep pace with technological advancements, the undercurrents of vitriol will continue to persist online.
Source link: Newsbreak.com.
