The festival of Eid al-Adha was celebrated all over the world, including India. Eid was celebrated on May 28th in India, while it was celebrated on May 27th in Arab countries and other countries. On the occasion of Eid, members of the Muslim community offer prayers (namaz) at mosques. A video claiming to be from London was widely shared on social media. In the video, a large number of Muslims can be seen sitting on the floor. Users have been sharing clips of what they say are Eid prayers being offered on the London Underground.
“London Underground Bakrid,” a user named Mahavir wrote while sharing the viral video.
link
Another user, Bhakt Prahlad, also shared the video and claimed it was from the London Underground. A user wrote the caption: “Britain has succumbed to Islamic extremism. Bakrid on the London Underground. This policy of appeasement will soon destroy Britain.”
link
Additionally, the video has been shared by several other users, including Concerned Citizen, Vivek Mishra, Subodh Sharma, and Roula Talj, who all claim to show Bakrid prayers being offered on the London Underground.
Fact check:
During the investigation, DFRAC Researchers have discovered that the viral video is not actual footage of prayers being offered on the London Underground. Instead, the video appears to have been created using artificial intelligence (AI). We first carefully examined the video and observed that the direction of the worshipers appeared to be inconsistent with Islamic prayer practices. Muslims do not pray in two rows facing each other. During congregational prayers, worshipers stand in parallel lines facing the Kaaba in Mecca, Saudi Arabia.
To further test this claim, our team searched Google using keywords related to the incident, including: “Bakrid Catfish on the London Underground” However, no reliable news outlets were able to find any reports about Eid prayers being performed on the London Underground.
I then analyzed the viral video using: AI-Detection tools. The video was tested at deep fake meterA platform developed by. University at Buffalo and UB Media Forensics Lab (UB MDFL). Several detection tools available on the platform indicated that the video was likely generated by AI. Here are the results: AVSRDD (2025): 99.9%AltFreezing (2023): 98.7%LIPINC (2024): 99.9%, LSDA (2024): 72.8%Tall (2023): 100%, XCLIP (2022): 86.6%
Meanwhile, our team analyzed a viral video on another site. A.I. Detection tool, Hive Moderation. This research also revealed: 94.9% Could be a viral video A.I.-generated. Additionally, to analyze the viral videos in more detail, our team made the following inquiries: A.I. Expert Kumar Aniket. After analyzing the video, Aniket provided some insights listed below.
1. Unusual similarities between facial and body structures
The video shows several people sitting side by side who appear to have very similar facial features, beard styles, hats, and body postures. Videos generated by AI often exhibit a “cloning effect” where subtle differences between individuals are reduced or lost.
2. Unnatural rendering of hands and fingers
In some frames, the hand position looks unusually static and artificial. Hands and fingers remain one of the most difficult functions for humans. A.I. Models often display unnatural proportions and stiffness when rendered accurately.
3. Subtle distortion of object geometry
There may be slight inconsistencies on the borders of subway poles, seat edges, and some prayer mats. AI-generated content often struggles to maintain consistent geometry and perspective across a scene.
4. Unusual visual composition
The overall scene looks very symmetrical and carefully arranged. The identical placement of the figures on both sides, the neatly aligned prayer mats, and the ideal camera angle create the impression of a staged composition. A.I.-Generated content.
Conclusion:
DFRAC An investigation has revealed that the video circulating on social media is not genuine footage from the London Underground. This video appears to have been created using artificial intelligence (AI). Therefore, the claims of the users sharing the video are false.
