Fact-Checked Lebanon Misinformation Was Mostly Real, Not AI
Most fact-checked misinformation in Lebanon’s 2026 conflict used real images or videos with false context, not AI-generated content.
Latest
AI Shows Different Sources Depending on Your Language
A study of ChatGPT Search, Gemini, Perplexity and Copilot finds that asking the same questions about Lebanon in English or Arabic can lead AI systems to cite very different sources.
Fatal Attacks in South Lebanon Got Less Global Coverage
Fatal attacks in South Lebanon and Nabatieh were substantially less likely to receive international English-language news coverage than attacks elsewhere in Lebanon.
Why the loudest voices online can look like the majority
Social media can make a small group look much larger than it is. This article explains why visibility, engagement, and public opinion are not the same.
Lebanon is building its digital state on fragmented data
Lebanon is investing in digital government, but disconnected data still limits how well institutions can understand fast-moving social and economic pressures.
Beirut Port Explosion Online Memory Study
A study of 7,191 X and Reddit records finds that online memory of the Beirut port explosion remains dominated by justice and state failure, suggesting the blast is still remembered as an unresolved political wound rather than a closed tragedy.
Uncertainty-Reply Asymmetry Replicates in English X Posts
A replication study finds that the uncertainty-reply asymmetry appears in English-language X posts across Federal Reserve policy, inflation, and electoral politics, suggesting uncertain language may invite more replies than certain statements.
Lebanon Crisis Communications Lacked Concrete Numbers
Analysis of 182 Lebanon crisis communications found that nearly 7 in 10 contained no concrete number. State/ministry sources were the least quantified, with only 6.5% including measurable figures.
13% of narratives drove repetition on X after ceasefire
A small share of narratives drove all repetition on X after the Lebanon ceasefire, concentrating repeated content within a limited set of clusters.
Media Coverage vs Public Demand in Lebanon (2026)
A data-driven comparison of media coverage and Google search behavior in Lebanon reveals a major gap between what news reports and what people seek.
Ceasefire posts drew 2.3× more engagement on X
Posts about the ceasefire attracted far more engagement than violation reports in the 48 hours after the Lebanon ceasefire.
35.5% of ceasefire posts repeated claims on X
Over a third of posts on X repeated existing claims after the Lebanon ceasefire, indicating that narratives spread faster than new reporting.
Data Briefs
30% of Arabic Tweets About Lebanon Contain Uncertainty
Nearly one-third of Arabic-language tweets about Lebanon contain linguistic uncertainty, showing that tentative language is a common feature of online discussions.
Data Briefs
Uncertain Tweets Get 25% More Engagement on X
Analysis of Arabic-language tweets shows that tweets expressing uncertainty receive higher engagement, particularly in replies, on X.