Disinformation under a networked authoritarian state: Saudi trolls’ credibility attacks against Jamal Khashoggi
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Abstract This paper deals with a case study that provides unique and original insight into social media credibility attacks against the Saudi journalist and activist, Jamal Khashoggi. To get the data, I searched all the state-run tweets sent by Arab trolls (78,274,588 in total), and I used Cedar, Canada’s supercomputer, to extract all the videos and images associated with references to Khashoggi. In addition, I searched Twitter’s full data archive to cross-examine some of the hashtag campaigns that were launched the day Khashoggi disappeared and afterwards. Finally, I used CrowdTangle to understand whether some of these hashtags were also used on Facebook and Instagram. I present here evidence that just a few hours after Khashoggi’s disappearance in the Saudi Consulate in Istanbul, Saudi trolls started a coordinated disinformation campaign against him to frame him as a terrorist, foreign agent for Qatar and Turkey, liar.... etc. The trolls also emphasized that the whole story of his disappearance and killing is a fabrication or a staged play orchestrated by Turkey and Qatar. The campaign also targeted his fiancée, Hatice Cengiz, alleging she was a spy, while later they cast doubt about her claims. Some of these campaigns were launched a few months after Khashoggi’s death. Theoretically, I argue that state-run disinformation campaigns need to incorporate the dimension of intended effect. In this case study, the goal is to tarnish the reputation and credibility of Khashoggi, even after he died, in an attempt to discredit his claims and political cause, influence different audiences especially the Saudi public, and potentially reduce sympathy towards him.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.007 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.003 |
| Science and technology studies | 0.003 | 0.001 |
| Scholarly communication | 0.005 | 0.042 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it