Digital Bullying and Hate Speech: Who Amplified Attacks on Egyptian Pilgrims?

A wave of hostile social-media posts targeting Egyptian Hajj and Umrah pilgrims began in January 2025, peaked during Ramadan in March and subsided in April and May. In this investigation, supported by Arabi Facts Hub (AFH) through its digital-tools support programme, Zawia3 examined whether the activity was spontaneous or coordinated, which accounts amplified it and how it related to the deaths during the previous Hajj season.

The analysis tracks posting patterns, timing, repeated wording and possible coordination. These can identify signs warranting further investigation; they do not by themselves establish who operated a campaign, whether accounts were automated or whether any state or organisation directed them.

Over the preceding months, X carried numerous Hajj-related hashtags, including “Hajj disaster,” “Hajj 1446 AH,” “Hajj is not safe,” “Pilgrims thank Saudi Arabia,” “Hajj is safe through God then the Saudi state,” “No Hajj without a permit,” “The Grand Mosque,” “The Sacred House of God,” “Saudi Arabia” and “Hajj season.”

Other recurring tags referred to “Hajj is a red line,” “We serve the sanctuary,” “The road to the rituals,” “Bushra hospitality,” “The guides,” “Hajj 2025,” tourism companies and their chamber, Mecca, rain in Mecca, the success of the 1445 Hajj season and “Their negligence killed the pilgrims.”

Alongside those messages, posts used nationality-related terms such as “Egypt,” “Egyptian,” “Egyptians” and derogatory variants, as well as “Enough,” “Kabsa and taameya together forever,” “Human invasion,” “Free,” and a tag demanding Egyptians’ deportation. Some posts targeted Saudi authorities or called for internationalisation of the holy sites.

The campaign followed the deaths in the 2024 Hajj, held from 14 to 19 June amid extreme heat. A WHO regional bulletin records Saudi authorities’ reported total of 1,301 deaths, 83% among unauthorised pilgrims. The Arabic investigation quoted a different total, 1,426 and 85%; the official figures are distinguished here. Reporting cited by the investigation put Egyptian deaths at 672.

Researchers searched terms including Egyptian Umrah pilgrims, Egypt and Hajj, Egyptian pilgrim deaths, pilgrimage problems, negligence and the Umrah crisis. They found more than 200 hashtags on X and examined activity from 1 January to 30 May 2025.

The most frequently recorded group combined “competent authorities,” “human invasion,” “free” and a truncated Egyptian-related tag, with 235 occurrences. “Two Eids” and “Citizens are the leadership’s priority” appeared 168 times. A group questioning Hajj safety or referring to a disaster appeared 137 times.

“We serve the sanctuary” occurred 97 times. Tags praising El Sisi and describing Egyptians as “a planet of joy” appeared 55 times. A group combining Hajj, an agreed plan and Egyptian nationality terms occurred 52 times.

“Enriching the guests” appeared approximately 35 times; tags about the Egyptian army’s readiness and public support for it, 34; “Egypt,” 24; “Great Saudi Arabia,” 22; Egyptian nationality variants, 18; and “Saudi,” 17. Variants of “Hajj is not safe” occurred ten times, the deportation-demand tag eight, and misspelled Egyptian nationality variants seven.

Mecca and Medina were repeatedly linked to pilgrimage and visits to the holy sites. Egyptian identity appeared prominently alongside allegations of exploitation, begging, disorder and misuse of meals intended for distribution. Negative emoji reinforced the tone. The likely context included claims about some Egyptian visitors’ conduct, meals and alleged begging or exploitation.

Negative sentiment dominated the sample

Meltwater analysis of 5,179 posts between 1 January and 30 May 2025 classified 3,133 as negative, 1,810 as neutral and 234 as positive. The source graphic also shows two unrated posts, completing the total.

March accounted for 2,458 negative posts—78.5% of all negative posts in the sample. January and February together recorded only 53.

The sudden rise coincided with Ramadan Umrah, followed by a decline over the next two months. The investigation treats the timing as a possible indicator of concentrated amplification or coordination. A peak around a major event can also reflect organic attention, so timing alone cannot establish a paid or automated campaign.

Negative content was associated with accounts located in 48 countries. Saudi Arabia led the reported negative count with 743 out of approximately 1,150 posts; Egypt recorded 229 out of 351; the United States 57 out of 90; the United Kingdom 25 out of 35; and Kuwait fourteen out of 22.

Other associated locations included the UAE, Bahrain, Iraq, Palestine, Russia, Qatar, Spain, Ghana, Sweden, Mauritania, Germany, France, Turkey, Jordan, Malaysia, Lebanon, Sudan, Libya, Syria, Morocco and Oman. Platform location data should not be read as verified nationality or proof of an account operator’s physical location.

A Python-based word-frequency analysis found the Arabic word for Umrah pilgrims 3,252 times and “Egyptians” 2,758 times. Other frequent terms included the sanctuary, tourism, Umrah, Mecca and food. Question words also recurred, reflecting inquiry, criticism or doubt.

The combination of repeated references to Egyptian pilgrims, food and tourism suggested a focused narrative. The investigation recorded 1,825 reposts among the 3,133 negative posts—58.3%. The Arabic article gives 61% in one passage and approximately 58% later; the percentage calculated from its stated counts is 58.3%.

Among the most active accounts in the negative-content sample, @_900_900 recorded 22 negative posts, @111104427l1ahnl seventeen, @cookingarba1 twelve, @abdullah1383_ eleven and @hamdamhamd3 ten.

Some posted no positive content in the collected sample. Their activity may reflect strongly held opinions or coordinated participation; these small individual counts alone do not identify bots or establish that these accounts produced most of the overall negative content.

More mentions, lower estimated reach

For the mentions-trend window, 1 January–26 May 2025, Meltwater recorded approximately 5,180 mentions, 553% above the previous period’s 793. The displayed daily average was 34, compared with five previously.

Activity rose sharply in mid-to-late March, with a daily peak approaching 800–900 mentions, followed by another smaller increase between 21 and 26 May. Early January to early March and much of mid-April to mid-May were comparatively quiet.

The March rise may be linked to Ramadan Umrah or a promotional or hostile campaign around it. The late-May increase could reflect renewed debate as the Hajj season approached. The chart alone does not determine the trigger or organiser.

The reach metric for the same window totalled 60 million, down 17% from 72.5 million in the previous period. Displayed daily average reach was 403,000, compared with 487,000.

Reach rose from approximately 22 March to 6 April, with separate peaks around 28 April–1 May, 6–11 May and a smaller late-May rise. It fluctuated more than mention volume.

More mentions alongside lower reach may indicate repeated discussion among smaller accounts or a narrower audience, while large accounts had greater influence in the comparison period. Reach is the platform tool’s metric, rather than a verified count of unique people who read the posts.

Repeated wording and a nationality-focused narrative

Initial analysis of negative posts found critical and accusatory language focused on Egyptian identity. References to food, the sanctuary and Mecca placed complaints in an emotionally charged religious and national context.

Common terms included Umrah pilgrims, Egyptians, tourism, food, sleeping outdoors, begging and theft. Other language evoked poverty, poor organisation, humiliation and loss of dignity. “Exploitation” and “money” conveyed implied accusations.

The investigation identified three recurring narrative components: assigning responsibility to tourism officials, ministers or the state; depicting distress through begging, sleeping outdoors and theft; and invoking Egyptian and religious identity.

Targets included Saudi authorities over conditions at the holy sites, Egyptian authorities or the Tourism Ministry over organisation and alleged failure, and Umrah companies and intermediaries blamed for exploitation.

Repeated references to nationality and alleged misconduct can turn criticism of particular arrangements into a broader hostile stereotype about Egyptian pilgrims.

Signs of content amplification

Heavy reposting, repeated links and recurring phrases across accounts suggested content circulation and reuse. The handle @alnaif_969 appeared 481 times, which may mean it was an amplification hub or an account repeatedly mentioned or targeted; frequency alone does not distinguish those roles.

The investigation highlighted similarities in posting levels, sentiment and vocabulary among @_900_900, @111104427l1ahnl, @cookingarba1 and @hamdamhamd3. It considered these possible signs of a directed influence network, while acknowledging that participants might be individuals expressing anger.

Several active accounts were associated with Saudi Arabia, others with Egypt or unknown locations. The report raised the possibility of cross-border coordination and a Saudi-centred network, but did not identify a controlling entity or provide independent evidence of paid operation.

The recurring allegations—begging, theft, bad food and disorder—could amplify a negative stereotype of Egyptians in Mecca rather than simply report individual incidents.

Python-based textual-similarity analysis found averages up to approximately 0.32 for selected accounts and substantial variation between comparisons. Similarity can indicate copied or shared wording. Without an identified model, baseline and further behavioural evidence, it does not establish common control or deliberate evasion of detection.

The source describes @_900_900 as especially active in negative content and relatively high in similarity to the wider discussion. Accounts associated with Egypt, Saudi Arabia and Bahrain indicate a regional conversation. Lower averages, such as the Egyptian account’s 0.093, may reflect more individual wording.

The different averages and maxima show that some posts shared language while others varied. These patterns are compatible with reuse of popular messages as well as organised amplification; they cannot conclusively separate the two.

What the patterns reveal—and what remains unknown

The investigation argues that human and religious concerns provided an entry point for politically accusatory or defamatory rhetoric. Egyptian authorities were blamed for poor organisation, Saudi authorities for treatment or management, and tourism companies and intermediaries for exploitation.

It identifies recurring accusations, national-identity framing, a concentrated March peak and substantial reposting as indicators of possible coordination. These patterns could arise from organised influence or from participants responding collectively to the same event.

The data show a concentrated wave of negative discourse targeting Egyptian pilgrims. They do not reveal who directed it, or establish that any named account belonged to a paid or automated network.

Although intensity declined in April and May, the vocabulary and stereotypes remained in pilgrimage-related debate. The report sees the wave as part of discourse that can return when a new Hajj season or crisis draws attention.

The findings raise questions about platforms’ role in amplifying nationality-based hostility and about methods for detecting coordinated activity without treating ordinary criticism as evidence of covert control.

Improved monitoring and accountable research are needed to understand the networks behind harmful campaigns and protect users from polarisation, particularly in debates shaped by religious practices and national identity.

This investigation was produced with support from Arabi Facts Hub (AFH).