More AI videos of small boat crossings circulate on Facebook

Full FactIndependentCenterFactual 98Objective 954 days ago

On 30 June 2026, Full Fact reported that several videos claiming to show members of the public discussing their views on immigration are AI-generated fakes. The videos, shared on Facebook, feature individuals speaking to an unseen person holding a microphone. These clips contain SynthID watermarks, indicating they were created using Google’s AI tools, along with visual anomalies such as buses passing through bollards and inconsistent background layouts. Full Fact identified these as AI fakes during its ongoing efforts to verify content on social media, emphasizing the importance of verifying sources before engaging with such material.

Bias read (Center): The article presents a factual assessment of AI-generated content related to immigration, a politically sensitive topic. While the subject is charged, the framing remains neutral, focusing on technical indicators (watermarks, visual glitches) rather than taking a partisan stance. The emphasis is on揭

Why these scores (Factual 98 · Objective 95): Highly accurate in citing SynthID watermarks as evidence. Visual glitches are noted as indicators. Slightly repetitive in structure but maintains neutrality.

Full FactIndependentCenterFactual 97Objective 9410 days ago

On 24 June 2026, Full Fact reported that several videos circulating on Facebook claiming to show illegal migrants arriving in the UK via small boats were AI-generated. These videos, which include scenes of men in life jackets on small boats alongside British border patrol vessels, were flagged using SynthID, a digital watermark from Google's AI tools. Multiple videos showed inconsistencies, such as misspelled terms like 'Burder force' and changing life jackets, indicating artificial creation. Full Fact confirmed that while hundreds of people have indeed crossed the English Channel by small boat recently, the videos in question are not authentic. The organization urged users to verify sources before sharing such content and provided resources for identifying AI-generated material.

Bias read (Center): The article presents a factual assessment of AI-generated content without overtly endorsing or criticizing specific political positions. It focuses on verifying the authenticity of media content rather than taking a partisan stance on migration policies or immigration debates.

Why these scores (Factual 97 · Objective 94): Accurately identifies SynthID in both audio and visuals. Describes specific examples of AI artifacts. Maintains a neutral tone though slightly promotional in subscription calls.

Full FactIndependentCenterFactual 96Objective 938 days ago

On June 26, 2026, Full Fact reported that a video circulating online depicting a large crowd in London protesting for 'halal standards' was identified as an AI-generated fake. The video, which included elements like a Google SynthID watermark, morphed faces, and garbled text, was shared on Facebook with claims of 'massive crowds shutting down London.' Full Fact confirmed the video was not authentic footage and noted similar trends of AI-created 'ragebait' content aimed at provoking outrage. They emphasized the importance of verifying sources before engaging with such content.

Bias read (Center): The article presents a factual assessment of an AI-generated video without overtly endorsing or criticizing specific political groups or ideologies. While the subject involves religious dietary practices and public perception, the framing remains neutral, focusing on technical verification rather on

Why these scores (Factual 96 · Objective 93): Correctly references SynthID and describes AI markers like morphed faces. Slight bias in phrasing 'massive crowds' but overall balanced.