Middle East Monitor / July 17, 2024
The Israeli Ministry of Diaspora Affairs is funding a multi-million-pound disinformation campaign spreading hatred of Muslims, Al-Jazeera has revealed.
The campaign, which targeted US lawmakers and spread Islamophobic content, began shortly after 7 October and has raised serious concerns about foreign influence operations and the use of artificial intelligence in spreading misinformation.
At the heart of the operation is Stoic, a Tel Aviv-based marketing company, which reportedly received $2 million from the Israeli ministry to run the campaign. The operation utilized AI to generate high volumes of deceptive online content, marking a new era in the ease and scale of influence campaigns.
The campaign’s existence came to light in early 2024 when researchers began noticing suspicious patterns across social media platforms. Scores of fake accounts were identified spreading pro-Israeli content, primarily focused on garnering support for Israel’s actions in Gaza. These accounts were predominantly linked to three websites: “Moral Alliance”, “Unfold Magazine” and “Non-Agenda”. Collectively, these sites amassed over 40,000 followers across Facebook, Instagram and X.
The campaign involved publishing pro-Israeli articles, often lifted from legitimate news outlets, and then sharing them through hundreds of fake social media accounts. One such example was an article reporting on allegations of UNRWA staff involvement in the 7 October attack. Researchers noted that the content was frequently repeated verbatim across multiple accounts, targeting the same individuals with identical posts and replies.
Perhaps most alarmingly, the campaign extended beyond merely promoting a pro-Israeli narrative. It is said to actively generate and disseminate Islamophobic content, attempting to mobilize far-right anti-Muslim sentiments. One website associated with the campaign, “United Citizens for Canada”, went as far as labelling Muslim immigrants a threat to Canada.
The campaign’s strategy appeared to be multifaceted, not only discouraging support for Palestinians but also attempting to drive a wedge between minority communities in the US and pro-Palestinian supporters. Experts following the campaign have argued that by actively promoting Islamophobia, the operation crossed the line from misinformation into hate speech, raising serious ethical and legal questions.
The scale and sophistication of the operation were made possible by the use of ChatGPT, an artificial intelligence tool created by OpenAI. AI was employed to generate most of the posts, allowing for a high volume of content to be produced and distributed rapidly. Both OpenAI and Meta (formerly Facebook) have since confirmed the existence of the influence campaign linked to Stoic in their respective reports.
In response to these findings, OpenAI and Meta have taken steps to prevent Stoic from using their platforms. However, experts question whether these measures are sufficient to combat such sophisticated disinformation campaigns. Critics argue that social media companies’ claims of effectively policing their platforms should be taken with a grain of salt, particularly as they increasingly rely on automated solutions while reducing human oversight.
The disinformation campaign is believed to be the first documented case of the Israeli government’s involvement in an online initiative targeting American and Canadian politicians, although the Israeli Ministry of Diaspora Affairs denies these allegations.
During the recent French election, Amichai Chikli, Israel’s diaspora affairs minister, endorsed the far-right anti-Sematic Marine Le Pen, in what analysts say is further evidence of the growing alliance between global Zionism and the global far-right.
