Meta Identifies AI-Generated Comments on Gaza Conflict

Meta found 'likely AI-generated' deceptive comments praising Israel's actions in Gaza.

  • Meta found ‘likely AI-generated’ deceptive comments praising Israel’s actions in Gaza.
  • The comments targeted users in the U.S. and Canada, linked to Tel Aviv-based firm STOIC.
  • This is Meta’s first acknowledgment of text-based generative AI in influence operations.

Meta has identified ‘likely AI-generated’ deceptive comments on Facebook and Instagram that praised Israel’s actions in the Gaza conflict. These comments appeared under articles from global news organizations and statements from U.S. lawmakers. The campaign mainly targeted users in the U.S. and Canada and was linked to Tel Aviv-based political marketing firm STOIC.

In its quarterly security report, Meta explained how deceptive accounts posed as Jewish students, African Americans, and other concerned citizens. This marks Meta’s first acknowledgment of text-based generative AI being used in influence operations. Although generative AI tools can quickly produce human-like text, Meta’s security team successfully dismantled the campaign and maintained their ability to detect and counter such influence networks.

Generative AI’s potential to spread disinformation has raised concerns among researchers. However, Meta’s head of threat investigations, Mike Dvilyanski, assured that the new AI technologies did not hinder their efforts to disrupt influence networks. Meta continues to monitor and address AI-generated disinformation to protect the integrity of its platforms.