The claim

A widely shared compilation appears to show London police repeatedly using vehicles or physical tackles to stop people riding scooters, with captions presenting the scenes as genuine action against thieves.

Reuters Fact Check concluded that the video is AI-generated. The short scenes contain physical and visual inconsistencies that do not behave like continuous real-world footage, and verification did not support the claimed incidents.

The verdict applies to the viral compilation, not to every crime report from London. Phone theft is a documented problem, and the Metropolitan Police have genuinely expanded an e-bike unit to pursue suspects in crowded urban areas.

Synthetic motion leaves clues

Generative video can produce convincing uniforms, streets and vehicles for a few seconds while losing consistency as objects move. Hands, wheels, body positions, shadows or lettering may change between frames without a physical explanation.

A dramatic collision can distract viewers from those defects. Replaying a clip slowly and checking whether equipment, faces and vehicle details remain stable often reveals problems that are easy to miss at normal speed.

Visual anomalies should prompt verification rather than stand alone as a universal detector. Compression, editing and motion blur can also distort authentic footage, so provenance and independent reporting remain essential.

The real e-bike program created a believable hook

Days before the fact check, Reuters documented the Metropolitan Police deployment of 20 high-speed e-bikes intended to pursue phone-theft suspects across terrain ordinary patrol cars cannot reach.

That real development makes fabricated action footage more believable. Disinformation frequently attaches synthetic images to a genuine event, allowing viewers to recognize the topic and assume the depiction must also be authentic.

Reuters reported that recorded phone-theft offenses fell from about 68,000 to 54,000 in the 12 months to May, while positive outcomes such as charges or cautions remained below one percent. Those figures provide context but do not validate the viral scenes.

How to verify dramatic police footage

Start with the earliest identifiable upload rather than the account that reposted it. Look for a location, date, original recording length and independent witnesses or local reporting tied to the same incident.

Check whether the police agency has released a statement or whether reputable news organizations can match buildings and road features. A caption that says only 'London' provides too little information for a sequence of supposedly serious collisions.

Content-credential systems can preserve information about how a file was captured or edited, but their absence does not automatically prove fabrication. Verification still requires several signals, not one badge or automated score.

Verdict: False

The viral compilation does not show genuine London police takedowns. It is AI-generated and should not be reposted as evidence of arrests, tactics or crime outcomes.

Real reporting confirms that the Metropolitan Police use e-bikes and other tools against phone theft. That authentic program is separate from the fabricated scenes.

The photograph with this article is a documented 2011 image of real Metropolitan Police officers and is visibly labeled as context. It is not a frame from the false compilation or a depiction of the alleged scooter incidents.

Limits and what to watch

The compilation is synthetic, but that finding does not mean phone theft is imaginary or that every video of London police activity is generated; each item requires its own provenance check.

The next development to watch is platform labeling, identification of the original generator account and any official response to continued redistribution.

Sources and further reading

Reuters Fact Check: Viral London police tackling-thieves compilation is AI-generated ↗

Reuters: Real Metropolitan Police e-bike program targeting phone theft ↗

Metropolitan Police: Official news and statements ↗

C2PA: Content provenance and authenticity standard ↗

Wikimedia Commons: Real Metropolitan Police photograph and license ↗