Syria in its "post-liberation" phase is living through unprecedented media transformations, as the digital space has expanded to become the "alternative arena" and the primary source of information. This openness, however, has been accompanied by a serious challenge: a flood of misleading content that no longer relies on traditional rumors alone and has come to be backed by artificial intelligence (AI) technologies, which now serve as a sophisticated instrument in "information wars."
With no unified media point of reference and with fake accounts multiplying, the Syrian public finds itself facing visual and textual material that looks highly realistic but is, in fact, a "digital illusion," either fabricated entirely or partially altered. Are we, then, facing a genuinely new weapon, or the same old rumor equipped with newer tools? This article attempts to approach an answer by first unpacking the concepts, then reviewing cases documented by fact-checking platforms since 2025, before turning to ways of confronting the problem.
A Turbulent Digital Space amid an Institutional Vacuum
With the page turned on the former regime, Syrians entered an era of sweeping informational openness. Social media platforms became the street's "thermometer" and the source of breaking news, at a time when traditional media institutions were losing their ability to keep pace with the flow of data and to debunk it in real time. Several actors have exploited this vacuum; with every pivotal event, whether tensions on the coast and in the south or a shift in regional relations, the disinformation machines start running again. What stands out here is a qualitative shift: an observable and growing reliance on artificial intelligence to generate images, videos and statements whose aim goes beyond sowing confusion to redirecting public opinion and creating a state of political and security disarray.
Unpacking the Term: What Do We Mean by Disinformation?
To understand the nature of this battle, we need to distinguish between the types of "information disorder" affecting the Syrian information space today, following the framework developed by researchers Claire Wardle and Hossein Derakhshan in a report they prepared for the Council of Europe in 2017:
- Misinformation: spreading false news by mistake, with no premeditated intent to cause harm.
- Disinformation: deliberately fabricating or spreading false information, with premeditation, in order to harm individuals, groups, institutions or states; this includes presenting genuine content in a false context.
- Malinformation: using genuine information with the intent to cause harm, often by moving what is meant to remain private into the public sphere, as with private leaks.
Syria has recently witnessed an intense mix of all three types, fed by open communication channels that lack mechanisms of verification and oversight.
Artificial Intelligence: The New Arsenal of Fakery
Fakery no longer requires editing experts, since generative AI tools have made enormous capabilities available to everyone for creating an alternative reality. Among the most prominent ways they work are the following:
- Visual generation: creating high-definition images and videos with tools such as Midjourney for images and Google's Veo for video, the latter able, since 2025, to generate clips complete with sound.
- Human simulation: running networks of fake accounts, some of them automated (bots), that impersonate local Syrian identities and interact and comment in a coordinated way; language models can give them the ability to write comments in convincing Syrian dialects.
- Document forgery: using language models to draft texts and statements that look like official decisions or government leaks.
- Informational hallucination: chatbots giving fabricated or false answers when asked about sensitive Syrian events, or when used to identify people in photos.
This last pattern seems closer to "misinformation" than to "disinformation," since it is not necessarily driven by premeditated intent; its impact, however, is no less serious when users treat a machine's answer as established fact. In April 2026, for instance, users relied on AI tools to "match" the features of a protester in Damascus with those of an officer accused of abuses, before the Kashaf team found that the result was entirely wrong and that the man was a former political prisoner.
Documented Cases: When Fabricated Images Deceived the Syrian Public
Monitoring by credible fact-checking platforms (Misbar, Verify-Sy and Kashaf) has documented cases in which these technologies, or digital editing tools more broadly, were used to create "false facts," including the following:
Fake Designs for the New Syrian Currency
In June 2025, months before the new currency was launched, Facebook pages circulated a design said to be the new 50-lira note that would be issued as part of the country's new visual identity. A Misbar fact-check concluded that the image was generated with artificial intelligence and that the Central Bank of Syria had not, at the time, announced any new denominations or designs.
Notably, this pattern did not stop after the new currency was officially launched on December 29, 2025, in six denominations, from 10 to 500 lira, bearing Syrian agricultural and animal motifs. Generated images followed the stages of the exchange, from a claim that a religious phrase was hidden in the 100-lira note to purported designs for a "5-lira" denomination that pages circulated in April and July 2026 and that Verify-Sy and Kashaf debunked. It may be that the listing of a 5-lira denomination in the decree on the new currency, while that denomination has yet to be put into circulation, lent these images a measure of plausibility.
A Satirical Image of a Bashar al-Assad Look-Alike Attributed to SHEIN
In September 2025, accounts on several platforms circulated an image allegedly used by the online fashion retailer SHEIN to display its products, showing a model who resembles Bashar al-Assad, along with a claim that the company had opened an investigation and apologized. Verify-Sy found that the image had been published by a satirical account that posts fake content, as a parody of an earlier controversy over an image of a model resembling the American Luigi Mangione on the same site, and rated the claim misleading. The fact-check does not indicate that the image was generated with artificial intelligence, which is a reminder that a fabricated image is not necessarily the product of these tools, and that satire, once stripped of its context, can turn into "news" circulated as fact.
A Generated Image Showing al-Sharaa Strolling through US Streets
An image circulated that allegedly showed President Ahmed al-Sharaa eating on a street in the United States, at the time of his participation in the UN General Assembly meetings in September 2025. A Kashaf fact-check found that it was generated with artificial intelligence, based on visual indicators, most notably a shop name in the background of the image spelled incorrectly ("Wafffl") and the complete absence of light reflections on the sunglasses, which does not usually happen in real photographs.
An Altered Image of Ahmed al-Sharaa's Meeting with Norway's Prime Minister
In the same period, an image circulated that its publishers claimed documented a "friendly" moment between President al-Sharaa and Norwegian Prime Minister Jonas Gahr Støre during the General Assembly meetings, with the two making a heart-shaped gesture. The Kashaf team found that the meeting did take place but that the image had been altered with artificial intelligence: the fingers appeared interlocked in a way inconsistent with the movement of the hands, with blurring around them and a difference in the distribution of light between the hands and the rest of the image, while the photos of the meeting available from news agencies showed no similar gesture.
A Fake Account Named "Mohammad Khair al-Sharaa"
In October 2025, the Kashaf team observed the spread of numerous accounts on social media bearing the "al-Sharaa" family name in an attempt to give followers the impression that they were close to the upper echelons of power. The team traced one of these accounts, named "Mohammad Khair al-Sharaa," and found that its profile picture originally belonged to the Egyptian doctor Mohamed Haroun Riad, who died in 2022, and that its owner had used AI tools to merge the doctor's face into composite images showing him with Hussein al-Sharaa, the Syrian president's father, and with other public figures.
A Fabricated Image Pairing Ahmed al-Sharaa with a Rapid Support Forces Fighter
In early November 2025, accounts promoted an image showing President Ahmed al-Sharaa with al-Fateh Abdullah Idris, a fighter in Sudan's Rapid Support Forces known as "Abu Lulu" and "the Butcher of El Fasher," who became notorious after those forces took control of the city of El Fasher in late October 2025. Verify-Sy concluded that the image was fabricated with artificial intelligence and generated from an old original photo of al-Sharaa with the journalist Taher al-Omar, a photo taken on the front lines of the Hama countryside and published in May 2019.
For all their variety, these cases reveal a traceable pattern: most of them revolve around the head of state and the state's symbols, and coincide with moments when public attention is concentrated, such as talk of a new currency, the president's participation in the UN meetings, or the events in El Fasher. They also fall along a wide spectrum, from generated images to partially altered real photos to satirical content taken out of its context. It appears that partially altered real photos are the most dangerous, because they borrow the credibility of the original event and then add a fabricated detail that the viewer finds hard to doubt.
2026: Fakery Takes More Forms
The phenomenon did not stop at the end of 2025. During 2026, the Kashaf platform documented cases in which AI fakery ranged across text, images, video and audio, the most prominent being:
- A document written by a bot (February 2026): accounts circulated a "government circular" banning eating in public during the month of Ramadan; the circular turned out to appear nowhere in the Official Gazette and to end with a leftover line from a chatbot's automated reply, in Syrian dialect: "If you want a shorter version or a more journalistic tone, let me know."
- A generated image of a religious site (April 2026): an image claiming to document an Israeli flag raised over the al-Frenj Synagogue in Damascus; comparing it with the actual building revealed missing stone steps and different windows and walls, while detection tools estimated a 99.9% probability that it was digitally generated.
- Fully generated video and audio (June 2026): a clip said to show Syrian army artillery drills near the Lebanese border, in which soldiers' bodies passed through parts of the gun and the Syrian flag appeared with two stars instead of three; the Hive Moderation tool estimated that both the video and its audio track were fully generated, with a probability above 98%.
- Partial editing of a real photo (September 2026): an image attributed to the recent fuel protests, showing an assault on a security officer, turned out to be an old photo from Ain al-Arab (Kobani) that had been altered with artificial intelligence.
Even so, the scale of the phenomenon should not be exaggerated. In August 2026, no more than five claims relied on generated or digitally altered images and clips, whereas 48 (42%) of the 114 claims that did not survive verification relied on genuine material that was old or filmed elsewhere, presented as current events, according to Kashaf's monitoring for that month, bearing in mind that these figures measure what reached the verification desk, not the actual volume of what circulated. This suggests that artificial intelligence remains an additional weapon in the arsenal rather than a replacement for its traditional tools, although its danger may lie in the nature of what it targets: the currency, the army, religious sites and the head of state.
How Do We Protect Society from "Manufactured Reality"?
Artificial intelligence poses a double danger in the Syrian context, since Syria's media environment is "soft" and its audience is emotionally charged and quick to be swayed by shocking news. A single generated image can spark a crisis or strike at public trust.
Confronting this requires action on three tracks:
- The role of professional media: filling the information vacuum quickly and accurately, and supporting fact-checking platforms (such as Misbar, Verify-Sy and Kashaf), which serve as the first "line of defense" by exposing fakery with technical evidence.
- Individual awareness: the biggest bet is on the user, because confronting disinformation begins with "healthy skepticism": checking the source, looking for the news on credible platforms, and resisting the urge to be first to publish. Patience is disinformation's greatest enemy.
- Looking to legal regulation: despite the difficulty of the current circumstances, looking to international experience is a necessity for building a Syrian system of digital protection in the future. On August 2, 2026, the European Union began applying Article 50 of its AI Act, which requires providers of generative AI systems to mark their outputs so that they can be detected automatically (with a grace period until December 2, 2026, for systems placed on the market before August 2), and requires anyone using these systems in a professional activity to disclose deepfake content. Added to this is the Digital Services Act (DSA), which lists the clear marking of this type of content among the measures that large platforms can adopt to mitigate the risks their services generate.
The cases above themselves offer practical guidance on what deserves scrutiny: hands and fingers, text and signs in the background of an image, reflections and the distribution of light, and details of flags and buildings compared with their originals, along with running reverse image searches to find the original version and checking what news agencies published about the same event. These signs, however, are likely to fade as the tools improve, and the result of automated detection tools is not enough on its own to establish an image's origin, which makes verifying the source and the context more important than staring at the details.
As for regulation, any future Syrian legislation needs to guard against sliding into the policing of expression. Law No. 20 of 2022 on combating cybercrime, inherited from the era of the former regime, is still in force, and since liberation its application has extended to prosecuting journalists and activists over posts on public affairs, despite the Ministry of Justice announcing that the law is under review, as documented by a Kashaf investigation. A transparency approach, based on labeling and disclosing generated content, may be better placed than a criminalization approach to protect the public.
To return to the question with which this article began: have AI tools become a new weapon in Syria's disinformation battles? The documented cases suggest that the answer is closer to "yes," with two caveats. The first is that this weapon, as far as the monitoring figures allow us to tell, remains less present than traditional tools such as recycling old images and clips; the second is that its effectiveness stems not from the tool alone but from the environment that receives it, one in which an authoritative source is absent and emotion outpaces verification. In other words, artificial intelligence did not invent disinformation in Syria; what it has most likely done is lower the cost of producing it and expand its persuasive power, especially when it targets sensitive symbols.
Understanding how these tools work and enabling the public to think critically are two indispensable steps for protecting the Syrian information space from turning into an arena of manipulation at this sensitive stage.
This article was first published in Arabic by Syria TV on November 28, 2025. This English version translates the revised edition published on Kashaf, which includes updates and additional cases documented by fact-checking platforms through September 2026, as well as corrections, most notably that the image shown in the first edition under the "Zara" case is a satirical image concerning the retailer SHEIN.
