A new investigation by the Arabic language fact checking outlet Kashaf identifies what researchers describe as an organized digital campaign supporting ISIS and attacking the Syrian government across several Meta platforms. The probe cites 3,841 Facebook posts tracked in recent months. According to Kashaf, 70% of those posts promoted ISIS, while the rest carried misleading political content aimed at undermining the post-Assad Syrian government. Kashaf reports many of the accounts behind the posts were created between June and October 2025, with activity rising during a recent government campaign against ISIS cells.
In its report, Kashaf asserts Meta’s content standards are “not consistently applied, particularly with regard to content from ISIS supporters, despite international recognition of the current Syrian government.” The findings raise concerns about how extremist actors continue to exploit gaps in the company’s enforcement systems.
Patterns of Bias and Overreach in Arabic Content
Kashaf’s investigation echoes long standing criticism from digital rights groups that say Meta disproportionately misflags Arabic language content while failing to apply its own rules evenly. The findings align with claims from Syrian journalist Qutaiba Yassin, who said on X that an employee at CCC, a company contracted by Meta, intentionally deleted posts documenting crimes by the former Assad regime. Yassin said the information came from inside the company’s offices.
Such claims deepen concerns that Meta’s enforcement has shown patterns of error, cultural misunderstanding and political bias when dealing with Arabic content. These patterns have also affected posts related to Syria’s revolution and armed groups tied to it.
SMEX, a digital rights group in the region, has argued that Meta has repeatedly removed documentation of developments following the collapse of the Assad regime in 2024. According to a July 2025 SMEX report, Meta took down posts featuring statements by then Hayat Tahrir al Sham (HTS) leaders as well as by Syria’s post-Assad interim president Ahmad al-Sharaa during the transition, prompting an Oversight Board review.
Meta said the removals were based on the Dangerous Organizations and Individuals policy and the Violence and Incitement policy. SMEX countered that both posts fit within policy exceptions allowing reporting on designated groups and conditional language directed at armed actors. “Meta’s moderation practices failed to adapt to the new political reality in time,” the organization wrote in its submission to the board.
Bad Actors Fill the Enforcement Vacuum
Researchers and analysts say these inconsistencies create openings for organized networks tied not only to ISIS but also to pro government and anti revolutionary actors. Kashaf’s report notes that some of the new accounts traced in its investigation appear to operate from neighboring countries. Syrian Information Minister Hamza al-Mustafa warned that roughly 300,000 fake accounts were spreading misleading content targeting Syrians. He said the Ministry of Information had documented the creation of up to 10,000 new accounts per day.
Experts say the overlap between Meta’s overenforcement against Syrian revolutionary content and its underenforcement against extremist or manipulative networks has allowed a broad spectrum of actors such as Iran, Hezbollah, ISIS and pro Assad loyalists to exploit the system more easily.
The growing body of evidence from Kashaf, SMEX and Syrian journalists alongside well documented systemic failings by groups like Human Rights Watch (HRW) suggests Meta’s approach to Arabic language enforcement remains fractured. As the company faces ongoing questions about political bias, transparency and cultural fluency, pressure is rising for clearer policies and consistent application in regions where online misinformation carries real world consequences.
