🔥 The Surge in AI-Fueled Misinformation

  The British Center for Countering Digital Hate (CCDH) has reported a staggering monthly 130% increase in disinformation on social network X, generated using artificial intelligence, over the past year. The study highlighted a significant rise in politically motivated deepfakes.

   This phenomenon was not analyzed on platforms like Facebook or TikTok, due to the general problem of lax moderation across social networks and the difficulties posed by the evolving capabilities of AI technologies. CCDH expects even more of such misleading content in the near future.
   In an effort to assess AI applications' safeguards, the researchers executed over 160,000 election-related tests. They succeeded in creating scenarios where President Joe Biden was hospitalized and fabricated brawls at various polling places.
  Of particular note, the image-generating AI Midjourney was found to create deepfake visuals in 65% of attempts, triple the occurrence rate in ChatGPT, despite having policies intended to prevent such misuse. Midjourney's CEO, David Holz, has vowed to remedy this in an upcoming update.

   It’s worth remembering that in January, AI was misused to clone Joe Biden's voice, dissuading citizens from participating in the Republican primaries, falsely claiming it would endorse Trump. This operation implicated two Texas firms, with the prosecuting attorney labeling the case as “genuinely unprecedented.”