Gaza and the rise of synthetic militarism: On genocide and generative AI

Abstract

Drawing on the case of the Israeli genocidal assault on Gaza, with attention to how Israelis and their supporters were employing synthetic images as political tools, this essay responds to current academic debates around visual generative AI in political contexts. The essay asks: How is the media landscape of war and political violence being reconfigured by generative AI? What is the political work of AI misinformation charges when they are directed by supporters of a violent state toward its victims? What are the longer colonial histories of such charges? And what are the effects of AI-generated synthetic realism in a context of a genocide? Engaging with and extending the scholarship on visual generative AI, this essay advances the concept of synthetic militarism – militarism by means of synthetic visual content and associated media practices – as a way of understanding of how AI-generated visuals are reshaping mediatized theaters of war and genocide across the globe.

Introduction

Today, in geopolitical theaters across the globe, the media landscape of war and political violence is being radically reconfigured by generative AI. While this proposition is all but a truism of the contemporary moment, it remains undertheorized, despite the rising interest in generative AI by media scholars, who have explored everything from creativity (Chateau et al., 2025; Laba, 2025), representation (Laba, 2024), ethics (Mager et al., 2025) governance (Obia, 2025) and the geopolitics of global cultures (Arora and Natale, 2025). At present, most scholarship on political effects of AI-generated visuals (i.e. images and videos) is structured around a central binarism: generative AI is perceived as a either a catalyst of misinformation (Gregory, 2023; Mahony and Chen, 2025; Twomey et al., 2023) or as an empowering new technology of synthetic realism and storytelling (Hausken, 2024; Schofield, 2023; Wasielewski, 2024). Taking the Israeli genocide in Gaza as our case study, this essay aims to move beyond such dyadic approaches by considering the ways AI-generated visuals are reshaping mediatized theaters of war. We propose the concept of synthetic militarism: militarism by means of synthetic visual content and associated media practices. Synthetic militarism, as this essay will demonstrate, rests on a dual ecosystem of both weaponization of misinformation detection and the embrace of synthetic realism for political purposes.

The analysis that follows builds on our theorization of “digital militarism” (Kuntsman and Stein, 2015). We coined this term over a decade ago, at the dawn of the social media age, to describe the weaponization of everyday digital communication tools, chiefly social media, into information warfare and patriotic support of a violent military state. In our earlier scholarship, which also focused on the Palestine/Israel case, the concept provided a means of studying the usage of digital tools by Israelis and their international supporters to advance the Israeli military occupation project, enabling an exploration of “the varied and often ordinary ways in which Israel’s military regime and pervasive culture of militarism are perpetuated and sustained” through ordinary social media practices (Kuntsman and Stein, 2015: 6). In what follows, we argue that the militarization of generative AI in the Gaza context emerges out of this technopolitical history.

While Israel’s genocidal assault is our case study, our argument extends well beyond this case. The first year of the genocide coincided with the introduction and rapid popularization of visual generative AI in political contexts across the globe, the earliest and most notable of which was Russia’s war on Ukraine (Laba, 2024; Twomey et al., 2023). The process would later expand to Gaza, Iran and other geopolitical contexts. Initially, the use of AI visuals in wartime social media posts was met with shock by the global publics and often accompanied by pronouncements of novelty (e.g. “the first AI war”; Davies et al., 2023). Promptly, novelty transformed into ubiquity. Today, AI-generated images of war are routinely sold on stock websites and synthetic visual content is deployed by state and military propagandists and visual activists alike. And while the response to such content varies greatly, online publics have come to expect it.

In what follows, we consider the following questions. How is the media landscape of war and political violence being reconfigured by generative AI? What is the political work of AI misinformation charges when they are directed by supporters of a violent state toward its victims? What are the longer colonial histories of such charges? And what are the effects of AI-generated synthetic realism, when used to depicting a future in which a genocide is over and the perpetrator has succeeded?

Building on over two decades of anti-colonial scholarship on the politics of media in Palestine/Israel (Aouragh, 2012; Azoulay, 2008; Hochberg, 2015; Nassar et al., 2022; Stein, 2021; Tawil-Souri and Matar, 2024; Yaqub, 2023), we aim to contribute to the growing body of work on the mediatization of the genocide (Chakrabarty et al, forthcoming; Della Ratta, 2025; Stein, forthcoming; Tawil-Souri, 2024). While much has been written about the unprecedented ways in which AI was integrated into the Israeli military arsenal after October 7, chiefly as a means of designating targets (Abraham, 2024; Goodfriend, 2023; McKernan and Davies, 2024), there has been far less scholarly attention paid to the militarized function of synthetic media. This essay explores the ways that synthetic militarism functions as both the grounds of a new warzone and a facilitator of genocidal politics.

Digital suspicion and its colonial histories

Our discussion begins with the first few months of the Israeli war on Gaza. The unprecedented scale of Israel’s genocidal violence was accompanied by an equally unprecedented volume of Palestinian-made visual evidence, circulated via global social networks. While Palestinians in Gaza had long employed social media to chronicle successive Israeli military assaults, the scale of such documentation and the degree of its global virality was new after October 7, registered by the recurrent trope of the “first genocide in real time” (Suny, 2024). ¹ This trope, so ubiquitous as to find its way to the floor of the ICJ in South Africa’s genocide case against Israel (Reuters, 2024), signaled both a deeply felt pain at the world’s inability to stop the unfolding events, and a hope within the Palestine solidarity movement that the unprecedented terms of digital witnessing might activate new forms of justice. For Israel and its global supporters, by contrast, the speed and scale of this Palestinian-made archive of genocide was a source of considerable concern, as it threatened the Israeli narrative of victimhood and self-defense. In the early months of the Israel’s bombardment, while the world was watching in horror, Israeli and pro-Israeli publics were focused on contesting the Palestinian-made visual evidence of wartime atrocities.

A campaign against what was perceived as the enemy’s use of AI was a pillar of this pro-Israeli project. The accusation was simple and easily replicated: namely, that Palestinian-generated image evidence was AI generated and thus falsified. The Anti-Defamation League (ADL) chronicled such accusations in a November 2023 memo on “Generative Artificial Intelligence and the Israel-Hamas War,” lamenting that “[d]isinformation peddlers are using GAI [generative AI] technology to tell a fictional story about the conflict. . . mostly glorifying Hamas, demonizing Israel and spreading already-debunked false narratives” (ADL, 2023). The Israeli press concurred, arguing that Hamas and its supporters were weaponizing AI “to try and display a false reality to the world. . . attempting to amplify the Palestinian tragedy worldwide” (Mann, 2023a). The accusation of AI falsification became a convenient means of dispelling with the political threat of Palestinian visibility.

The pro-Israeli campaign against Palestinian weapons of synthetic misinformation would quickly gather steam. Israeli professionals from the commercial high-tech sector were called upon to lend their expertise to this wartime effort, to help sort fact from wartime fiction, even as the value of their start-ups soared (Lebovic, 2023). They argued that while wartime misinformation itself was not new, the “emerging deepfake cyberfront” was unprecedented: “A dramatic super-evolution of misinformation is taking us by storm. . . We are at the start of a deepfake revolution” (Lebovic, 2023). AI detection guides began mushrooming in both the Israeli and pro-Israeli media, in an effort to school media consumers in the micro-practices of fraudulence detection and thus protect Israel from a “flood” of falsified AI-generated evidence (Mann, 2023b). Many such detection guides quoted from digital experts in image verification, or CEOs from the AI private sector, listing recognizable signs of inauthentic images and offering tools for laymen online detectives who could jump into the breach of evidential crisis, where the Palestinians and their supporters were the seen as the chief offenders. Beware of fingers that look too uniform, they advised, wrists that are not bent enough, and at the wrong angle, feet that are too large or toes too perfect. Pay attention to patterns that are blurred, lighting that seems staged (Mann, 2023a). This work of detection, many experts argued, should be routinely performed by social media users to repudiate the volume of fake photographs and videos coming from Gaza (Klein, 2023). Such acts of detection were framed as practices in patriotism, vital means of supporting Israel in its existential struggle.

In tandem, one saw the emergence of a burgeoning industry of “fact checking” platforms that began offering services to help navigate the world of proliferating synthetic images. At the dawn of generative AI, such platforms were spreading rapidly in numerous media contexts across the globe. But in the genocide context, they were being conscripted into the Israeli hasbara (propaganda) efforts against the veracity of Palestinian image-evidence. Israel advocacy organizations invested heavily in such guides (e.g. the ADL and Aish; ADL, 2023). ² Their work was deemed a crucial means of “address[ing] heinous abuses of GAI [stat] technology,” all in the service of protecting Israel’s international image in the court of public opinion (ADL, 2023). In the process, the burgeoning field of amateur digital forensics, grounded in open-source intelligence (OSINT), was being harnessed to the pro-Israel project of dispelling the threat of the Palestinian online archive (Mallinder, 2024). Digital experts and self-styled OSINT investigators framed the moment as an urgent battle over AI “denial and distortion” (ADL, 2024).

In some sense, none of this was new. Just as Israeli genocidal violence in Gaza has been deeply implicated in the much longer legacy of Israeli settler colonialism and military occupation, the same can be said of pro-Israeli suspicion of Palestinian image evidence. Although such suspicion took a new shape in the genocide’s early months, harnessed to a refutation of the enemy’s synthetic image, it was rooted in the historical tenets of Zionist colonial denial: namely, denial of Palestinian history, national identity, land claims and humanity itself (Said, 1981). For its part, pro-Israeli suspicion directed against the Palestinian digital image has a more recent history, as traced in our earlier work. We have argued that during the first decade of the social media age, the theology of denial would be tethered to emerging social media tools and practices, resulting in a technopolitical formation we called “digital suspicion”: namely, a mode of suspicion directed against Palestinian digital images and archives of Israeli state violence (Kuntsman and Stein, 2015). This term aimed to describe the ways in which pro-Israeli publics, at the dawn of the social media age, were impugning the veracity of the growing archive of Palestinian eye-witness evidence that was consolidating on the internet in the first and second decades of the 21st century. Through digital suspicion, this evidence was reframed as staged or digitally doctored and thus fake.

These earlier contexts and examples of digital suspicion were varied, but all were grounded on the same base claim: Palestinians depicted in digital images were not really dead or injured. Rather, it was proposed, such images had been maliciously created to malign Israel in the eyes of the international community, and it was therefore a patriotic duty of Israeli and pro-Israeli publics to dispel these baseless accusations. In each campaign, digital suspicion by means of amateur digital forensics was portrayed as the ultimate act of good citizenship in a wartime context, a means of defending the Israeli state in a time of existential threat. By claiming that images were not real, Israeli patriots and their supporters endeavored to semiotically refigure the growing online visual field of Israeli violence as a means of exonerating Israelis and discrediting Palestinians. According to this logic, if the Palestinian dead were not real but digitally manufactured, then Israeli culpability, too, was a manufactured fiction.

Pro-Israeli digital suspicion, we are arguing, did not simply emerge in 2023 with the entrance of generative AI tools and platforms. Rather, this technopolitical phenomenon had been steadily growing and morphing for more than two decades, keeping pace with both changing political fields and shifts in technological advances and aptitudes. In the early 2020s, the pro-Israeli art of digital suspicion had already developed into a near industry. By 2023, amidst the genocidal campaign in Gaza, it had become both rampant and highly structured, consolidating into a form of high-tech colonial nationalism that drew on the expertise of a cybersecurity army of laymen internet experts and OSINT investigators. The newer grammar of suspicion shared much with its earlier forms, down to the proliferation of amateur digital forensics, conscripted to detect falsehood by directing online eyes to incongruous perspectives, suspicious angles, too many fingers or wrists bent in the wrong direction. But in the genocide context, it had an additional chilling effect. Palestinians in the images were subjected to a double eradication: both as targets of military destruction, and as objects of forensic digital investigation designed to invalidate.

In these campaigns over image veracity, AI functioned as a site of political proxy: a technopolitical and discursive locus where larger political debates were reframed as mere contestations over image veracity. The focus of amateur digital experts-cum-patriots on alleged AI misinformation served as a distraction from the real battleground: the ongoing carnage waged by Israel on Gaza. The battle against the synthetic image functioned as an alibi, a means of contending with the threat posed by the “genocide in real time” by persuading global audiences that the events they see on their mobile screens were mere pro-Palestinian distortions aimed at discrediting the “real” victim: Israel.

Perpetrator fabulation and its genocidal imaginaries

While digital suspicion campaigns were at the forefront of synthetic militarism in the first few months of the genocide, another mode of pro-Israeli engagement with AI was also gaining ground. At work was a gradual embrace of AI tools to advance the state’s genocidal agenda. The early forms of such embrace varied. Advocacy or hasbara usages predominated in the early stage, including text-based tools such as “Hasbara GPT,^” which generated answers about ongoing political events in Gaza from the Israeli advocacy standpoint (YESCHAT, n.d); the “Speak Up for Israel” app, an AI powered “world class persuasion expert” (Reasonate, n.d.) and AI chat bots that argued online with critics of Israel (Buxbaum, 2024). At around the same time, Israeli and pro-Israeli users were also experimenting with the use of synthetic images to imagine post-war futures that aligned with Israeli state priorities.

The latter is the focus of our discussion here. We call it perpetrator fabulation: a mobilization of generative AI technologies as a tactical tool of pro-Israeli information warfare and political imagination. Perpetrator fabulation emerged alongside digital suspicion, but with an inverse logic: instead of fighting the AI misinformation battle as a mode of patriotic engagement, perpetrator fabulation rested on the embrace of synthetic images as part of a settler-colonial war of imagination. It began in an ad-hoc manner on the margins of the pro-Israeli response to the war and was initially far outpaced by the volume of digital suspicion directed against the ostensible AI images of the enemy. As the genocide progressed, perpetrator fabulation would gather steam, increasingly normalized as a mode of digital patriotism, intended to support the Israeli military assault.

Consider, for example, a viral set of images created by a Hebrew-language Facebook user with a distinctly Israeli name, shared widely on social networking platforms and reposted multiple times in global news, social media outlets and numerous fact-checking websites (Africa, 2023; Alam, 2023; Kulsum, 2023; Reuters Fact Check, 2023). The original Facebook post contained a series of manifestly synthetic renderings of a future Israeli victory, including soldiers and civilians wrapped in prayer shawls or Israeli flags and an abundance of celebrating crowds in ostensibly Israeli urban spaces – all embracing the smooth esthetic of AI imagery. The post proudly declared its usage of AI as a tool to imagine – and create – a different political reality:

A lot of this war is psychological so let's imagine our victory, and the victory will come! To clarify: of course in the war there are many casualties however there is a “but”. We must keep high energy and positive attitude. We can't let what happens break our spirit. Look at the positive side of the picture! The images were created using artificial intelligence. (Shefi, 2023, emphasis ours)

In this celebratory embrace of the synthetic image, AI was employed to cement the national sentiment of the moment: namely, the overwhelming Israeli support for the military offensive on Gaza. Such images simultaneously erased the horrific human cost of the offensive and proactively conjured up a desired Israeli future.

A different modality of perpetrator fabulation was mounted in the streets of Tel Aviv in the early week of Israeli bombardment. A billboard campaign, backed by a private Israeli organization, was (in the words of the Israeli press) “created using AI, feature[ing] images of the leaders of Hamas and Hezbollah. . .as [if] . . .beaten, defeated, and captured by the fighters of Israel's security forces.” The Israeli press continued:

These are the first billboards in Israel to be made with AI technology. The billboards have three main purposes - to restore the Israeli public's confidence in the ability of the security forces to keep citizens safe, to remind the security forces and leadership of the importance of reassuring the people, and to restore the deterrence power of the State of Israel in the face of its enemies (Borden, 2023).

In this example, celebrated by the Israeli media in the language of technological pioneering (“the first billboards. . . to be made with AI”), the military violence that was expunged from the Facebook images was moved to the foreground, but in highly selective form: as violence inflicted on Hamas by the Israeli military. The horrific impact of Israeli attacks on Gazan civilians was invisibilized. In this instance, an AI-generated future of Israeli victory migrated from social media into the streets of Tel Aviv in the form of massive billboards, placed at prominent urban locations. In the process, generative AI literally functioned as a technology of world building: a means of remaking Israeli public space in the image of the state’s dominant political imagination.

Our final example of perpetrator fabulation is a video advertisement that aired on Hulu in the winter of 2024. Although Israeli state bodies never confirmed the ad’s provenance, the clues were multiple: the ad ended with the emblem of the state of Israel (Fox, 2024; Graham, 2024; Hoover, 2024) and the clip was hosted on the webpage of the Israeli Public Diplomacy Directorate, under the Hebrew slogan “together we will win” (National Information System – Together We Will Win! [in Hebrew], 2024). Employing the recognizable genre of the tourist advertisement, the video used AI to reimagine Gaza as a highly desirable leisure destination replete with immaculate beaches, five-star hotels and upscale restaurants, teeming with tourists. The 39 second video ended abruptly with scenes of Gaza devastated by the ongoing war. “This is how Gaza would have looked like,” said the narrator, “without Hamas.” The message itself was unremarkable: echoing well-worn hasbara talking points, it blamed wartime destruction on Hamas rather than on Israel. What was different was the set of visual and epistemological strategies employed. Here, the Israeli state or its proxies openly embraced generative AI as a way to narrate Israel’s wartime ambitions and depict an alternative future for Gaza.

In this rendition, which mixed photorealism with the sleek visual language of generative AI, the synthetic image acted as a veil: a cleansing tool that disappeared all signs of Gazan devastation inflicted by Israel, leaving only a prosperous generic tourist scape. This synthetic story of Gaza offered a split temporality, simultaneously oriented toward an imaginary Gazan past (“this is how Gaza could have been”) – a past stripped of Palestinian history and resistance to the Israeli occupation – while also presenting an imaginary future image of what Gaza could/should look like under Israeli control. But this fictional future Gaza was not just a synthetic vision. Rather, it gave visual form to Israeli political fantasies of a Gaza without Hamas, remade as tourist site or a real estate opportunity. Here, AI functioned as a vehicle of militant colonial ideology, conscripting the generic smooth visuals of generative AI into an expulsionist project. The timing was chilling: the sleek video was produced amidst horrific and escalating genocide.

These forms of perpetrator fabulation, albeit different in medium (a Facebook post, a billboard image, a video), were all part of the same strategy. They were conscripting synthetic images into violent genocidal future-making, a process Donatella Della Ratta has poignantly described as “speculative violence” (Della Ratta, 2025). Generative AI, here, offered a way to fabulate the political future that most Israeli sought but could not formulate in standard political grammars, rendered through the smooth and generic esthetics of the synthetic image.

Conclusion: Synthetic militarism and its futures

This essay was driven by several questions pertaining to the militarized use of visual generative AI in political contexts. We asked: how is the spread and normalization of synthetic images changing the political terms, formations and effects of war and violence? What kind of political work is being performed through such images and alternately, the repudiation of them? And, in the case we have studied here, what was the role of such visual content in aiding and abetting the Israeli genocidal and colonial project?

Over the course of this essay, we traced two contrasting pro-Israeli tactics toward the synthetic image as they took shape in the first few months of the genocide. First, we chronicled the pro-Israeli outcry about an epidemic of Palestinian AI misinformation, a phenomenon that we read as an intensification of one of the key principles of digital militarism: digital suspicion. Second, we tracked the ways that Israelis and their supporters were embracing synthetic images as technologies of future world making. We have called this perpetrator fabulation: a means of crafting synthetic realities to visualize the political outcomes that Israeli sought but could not formulate in standard political grammars.

The onset of perpetrator fabulation in the early months of the Israeli assault on Gaza marked a fundamental change in the pro-Israeli engagement with generative AI. While digital suspicion would remain a site of patriotic political engagement, it would be gradually overtaken by the proud embrace of perpetrator fabulation, a paradigm shift in synthetic militarism whose contours would only become perceptible over the course of the ensuing year and a half of the genocide. At the time of this writing, these contrasting modes of engagement with visual generative AI coexist in the mediatized theater of politics in the Gaza-Israeli context, just as they do globally. Together, they form the dual ecosystem of synthetic militarism, which rests on both the weaponization of truth-claims, and on the embrace of synthetic realism for political purposes. Both modalities have aided in the invisibilization of Israeli-inflicted violence on Gaza, and therein Israeli accountability for such violence. If digital suspicion has extended the colonial legacy of banishing perpetrator violence from the visual frame, perpetrator fabulation has worked to unapologetically assert and bolster such violence through a beautified, smooth AI esthetic. With its dual ecosystem, synthetic militarism moves well beyond current scholarly debates around AI-generated images, often conceptualized around the dyad of misinformation versus synthetic realism, with which this essay began. Synthetic militarism, by contrast, traffics in both.

The gradual normalization of perpetrator fabulation in the second year of the genocide coincided with a transitional moment in Israeli politics. In our earlier work we argued that Israeli colonial nationalism, as it took shape in the first decade and a half of 21st century, was organized around the principle of “public secrecy” (Taussig, 1999): a collective agreement, albeit always fragile, to know what not to know. In Israel during this period, public secrecy took the form of living with military occupation and colonial complicity by colluding in the illusion of their absence. We argued that emerging forms of digital militarism worked to preserve and protect the fragile terms of such secrecy. Much would change in the ensuing decade, as Israeli society moved ever rightward, amidst the growing and open embrace of supremacist settler ideologies. In turn, the fragile realm of public secrecy was eroding.

The genocidal attack on Gaza marked the most spectacular dissolution of public secrecy, taking shape in the full-throated Israeli embrace of the destruction of Gaza and perpetrator fabulation was a powerful symbolic locus of this articulation. By the winter of 2025, a year and a half into the genocide, the AI-enabled fantasy of a Gaza without Gazans had been normalized, perhaps most clearly enunciated through the infamous “Trump Gaza” video (Holmes and Owen, 2025). The viral video depicted Gazans walking from their war-devasted ruins into a sun-lit “Riveria” of skyscrapers, pristine beaches, tourist-filled streets and luxurious hotels. A few months later, the Israeli Science and Technology Minister Gila Gamliel posted a similar AI-generated video on X (formerly Twitter). In a clear homage to “Trump Gaza,” it depicted Gaza as a thriving techno-futuristic resort, cleansed of both the signs of Israeli military carnage and of Palestinians themselves. “This is what Gaza will look like in the future,” Gamliel wrote in an accompanying caption. “Voluntary emigration of Gazans – only with Trump and Netanyahu. It’s us or them” (JTA and ToI Staff, 2025). The videos concretized what earlier Israeli synthetic images had more modestly imagined: bringing the full genocidal future into being. In the hands of Israelis and their supporters, visual generative AI functioned as a powerful colonial tool, fashioning a political future from which Israeli state violence was at once invisiblized, completed and erased from national memory.

Does generative AI lie at the root of Israel’s horrific violence inflicted on Gaza? Certainly not, although Israel’s adoption of AI for wartime targeting and large-scale automated surveillance has been crucial in this violent project (Abraham, 2024). But has synthetic militarism strengthened Israeli and international delusion whereby the ruins of Gaza can magically disappear into the “beautiful future” that “awaits them” (Sky News, 2025)? Definitely. Amidst the rapid expansion of both generative AI and Israeli coloniality, the power and potential of synthetic militarism is advancing in ways that we can only begin to imagine.