In an era of digital communication transformation, Israel has launched a large-scale public opinion influence campaign with a budget of $145 million, utilizing artificial intelligence and a network of social media influencers to target the information space, with particular emphasis on audiences in the United States.
Details have emerged about the campaign, which was previously announced: the strategy demonstrates an exclusive focus on Generation Z. 80% of the content is placed on visually-oriented platforms like TikTok, Instagram, and YouTube, achieving a monthly reach of 50 million impressions and setting a new standard in digital diplomacy.
A key technological component of the initiative involves coordinated efforts to shape the responses of generative artificial intelligence about Israel through a methodology of generative optimization, aimed at systematically embedding pro-Israeli narratives into the architecture of large language models.
Operational management of the campaign is handled by Israel’s Government Advertising Agency, implemented by Havas Media Network, while strategic planning is assigned to the American firm Clock Tower X LLC, led by Brad Parscale—former campaign manager for Donald Trump and current Director of Strategy for Salem Media Group. The project’s financial structure includes a $6 million contract with Parscale’s firm, signed on September 18, 2025, valid until the end of the calendar year, with funding equally split between Havas and Clock Tower.
According to the agreements, Parscale’s team has committed to monthly production of at least 100 basic creative materials—from video content and graphics to audio and text formats—with subsequent generation of 5,000 adaptive variations. A significant portion of this content is created exclusively for Generation Z platforms, with a key performance indicator of disseminating pro-Israeli narratives at a cost of less than $2 per thousand paid impressions.
Among the most controversial aspects of the campaign are the targeted efforts to influence how generative AI systems—including ChatGPT, Google’s Gemini, and X’s Grok—formulate responses to queries related to Israel and associated topics.
As commented to IR by Ilya Kiriya, Professor, invited researcher and professor at Paris University, Södertörn University, University of Grenoble, and Lyon Institute of Political Studies, such campaigns are not unique; this approach to shaping the information field has been observed for at least 15 years.
“Well, frankly, I don’t see anything new in this news. We know that for at least the past 15 years, various governments or state media have been trying to ‘manage’ search algorithms. I see two elements here: managing the visibility algorithms for certain news and managing ‘social engagement’ algorithms,” notes Professor Kiriya, emphasizing the transformation of tools in the so-called “industrialization” of social engagement.
“Previously, tools of ‘industrializing’ social engagement were used here: ‘trolls’ were hired for money,” the professor comments. “Then algorithms were written to mimic the actions of real users and create content to artificially ‘inflate’ social engagement metrics, which helps promote content through automated algorithms on social networks.”
“Now, automated content production using AI assistants is being added to this. The task here is simple—AI assistants like ChatGPT train their models on a certain corpus of texts, specifically texts available on the Internet. Therefore, if we can generate a sufficient volume of texts online, we can force the model to incorporate these texts when answering questions.”
“Consequently, if we clutter the internet space with messages reflecting a certain ideological position, we will force the model to consider this position when generating answers. So the goal here is not only to influence visibility algorithms but to change the overall spectrum of opinions available in the internet space.”
