Jean Baudrillard’s idea that the Gulf War “didn’t really happen” wasn’t to say that people weren’t dying in Iraq, but rather that what the public witnessed was a carefully curated TV show. It was a production that cleaned up the messy reality of war, detaching it from the profound suffering it caused. For Baudrillard, the way the war was shown became the event itself, swallowed up by its own filtered image. The actual fighting simply provided raw material for this media spectacle.
Now, almost 35 years later, with the rise of generative AI and a world constantly embroiled in conflict, Baudrillard’s insights feel even more relevant and urgent. Television started this blurring of lines, and today, technologies like large language models and Deepfakes are scaling it to an astonishing degree. Information isn’t just passively relayed anymore, with biases subtly influencing what’s shown. Instead, our entire information landscape is under threat, especially with everyone glued to their smartphones. We’re seeing whole events being made up from scratch, and “evidence” being manufactured to support them. Researchers like Shirin Anlen and Mahsa Alimardani have aptly named this new era “forensic cosplay.” This is where even the tools designed to spot fakes are now being used to create convincing illusions. Imagine fake heatmaps or scientific-looking visuals giving a stamp of authority to conclusions that were decided long ago. They shared an example where a viral claim about a New York Times photo from Tehran being AI-generated got over 600,000 views, even though the analysis was done on a screenshot of an Instagram post, not the original image. And as often happens, by the time fact-checkers weigh in, the damage is already done; the misinformation has spread far and wide, making a lasting impression that truth struggles to undo.
The truly alarming part isn’t just that false information – propaganda – is spreading, which has always been a wartime tactic. It’s how it’s spreading: through systems built to maximize clicks and shares, not truth. Many people, having lost faith in traditional news, turn to social media for quick updates. These platforms aren’t innocent bystanders; their very design often fuels the fire. Their recommendation algorithms, autoplay videos, and engagement-focused feeds are simply not built to tell the difference between a verified news report and a Deepfake crafted to spark outrage. Even worse, these Deepfakes are often helped by the algorithms because they’re designed specifically to trigger strong emotions, leading to more shares, comments, and endless replays – all of which pads the platforms’ pockets.
Studies, including over 30 articles published by Tech Policy Press between 2021 and 2026, consistently show that in conflict zones from Gaza to Ukraine to Iran, the way we get information has become a battleground itself. Social media platforms, consciously or carelessly, act as stage managers. Nusrat Farooq’s analysis in July 2024 pointed out that generative AI has removed the need for special language skills or technical know-how to create influence operations, making it easier for anyone to spread disinformation. Both the Stanford Internet Observatory and Georgetown CSET agree: there’s no easy technical fix for disinformation generated by large language models. Yet, platforms have paradoxically dismantled their own defenses. Over the last three years, trust and safety teams — the people who deal with harmful content — have been cut or even eliminated. Even labels identifying state media have been removed. This isn’t accidental; political forces, particularly right-wing interests in the US, have cleverly rebranded content moderation as “censorship,” benefitting directly from the resulting information chaos.
This creates a dangerous cycle, as Prithvi Iyer explains, drawing on a WITNESS report from September 2024. First, there’s “plausible deniability,” where real evidence can now be easily dismissed as AI-generated. Second, there’s “plausible believability,” meaning synthetic content that confirms what people already believe is accepted without question. These two forces don’t just muddy the waters; they erode the very foundations of how we understand the world and have meaningful conversations in a democracy. If everything could be fake, then nothing needs to be believed. This leads citizens to abandon trusted institutions and retreat into online echo chambers, where algorithms feed them only what they already want to hear, reinforcing their existing views.
This isn’t just a problem for one conflict; it’s a global, systemic issue. But its manifestation in India carries particular dangers for democracy. The Bulletin of the Atomic Scientists has explicitly warned that Deepfakes during India-Pakistan crises could lead to “catastrophic misperception and miscalculation” between two nuclear powers. Imagine a fake video of a military chief admitting defeat goes viral, reaching hundreds of thousands before it’s debunked. The tiny window for de-escalation, before public fury, political pressure, or military blunders trigger something terrible, is terrifyingly small. Furthermore, modern conflicts often become an excuse for widespread censorship. During India’s “Operation Sindoor,” over 1,400 URLs were blocked. We’ve seen this pattern recently with satirical posts about the Prime Minister being blocked on social media, and internet shutdowns imposed across Jammu and Kashmir. When facing information disorder, the state’s go-to reaction isn’t to promote media literacy or support fact-checking, but to reach for the blunt instrument of a blackout. This backfires, as those cut off from official news ironically turn to rumors and misinformation for updates.
So, what would a serious solution look like? First, we need to hold platforms legally accountable when their algorithms amplify synthetic content during active conflicts. This isn’t about a blanket ban, but a targeted expectation: platforms must actively reduce the algorithmic spread of unverified conflict content, with a duty of care, not a duty of censorship. Second, there needs to be significant public investment in verification infrastructure. This means robust media literacy programs, easy-to-use open-source tools for detecting fakes, and strong support for independent fact-checking organizations, which are currently outmatched and underfunded. Third, we need international legal frameworks that treat the weaponization of information during armed conflict as a humanitarian issue, not just a matter of platform rules.
Baudrillard spoke of simulation replacing reality. AI has taken us further, to a point where reality and simulation are technically indistinguishable. The platforms that mediate both have no commercial incentive to help us tell them apart. Social media, fueled by generative AI, is like a multi-pronged attack on our shared understanding of truth. The “simulation” isn’t created by a few gatekeepers anymore; it’s generated by everyone, for everyone, optimized by algorithms that value engagement above all else. We must realize that every war now is Baudrillard’s Gulf War, but with an even darker twist: now, no one even pretends to look for the truth anymore.