Fake AI-generated image of damaged bridge causes rail disruption
Rail services were disrupted after a suspected AI-generated image appeared to show major damage to a bridge.
Listen to this article
The image - which showed that parts of Carlisle Bridge in Lancaster had collapsed - circulated on social media after a 3.3-magnitude earthquake hit north-west England on December 3.
Network Rail said it was made aware of the image at around 12.30am in the early hours of Thursday morning and stopped rail services across the bridge while safety inspections were carried out.
No damage was reported after checks were made. The railway line was fully reopened at around 2am.
32 services - including passenger and freight trains - were delayed because of the image.
Network Rail added that most of the delays from the incident were local but due to the length of the West Coast Main Line disruption was felt as far north as Scotland.
A spokesperson also urged people to think about the "serious impact" of creating or sharing hoax images.
'Completely unnecessary delay'
A spokesperson for Network Rail said: "The safety of rail passengers and staff is our number one priority and we will always take any safety concerns seriously.
"The disruption caused by the creation and sharing of hoax images and videos like this creates a completely unnecessary delay to passengers at a cost to the taxpayer.
"It also adds to the high workload of our front line teams, who work extremely hard to keep the railway running smoothly.
"Please think about the serious impact it could have before you create or share hoax images."
Read more: Ex footballer Joey Barton avoids jail over ‘grossly offensive’ social media posts
Read more: Law student locked up in Dubai prison for 25 years on cocaine charges to be freed
'Spreading like wildfire'
LBC's Technology Correspondent, Will Guyatt said: "We're already at the stage where it's as easy to create a fake image on your smartphone as it is to call a member of your family if you know what you're doing.
"What required a lot of effort a year ago is now as simple as typing in a description into an app and the AI does the rest.
"It's getting harder and harder for the trained eye to distinguish between real and fake images and it's only going to get tougher.
"There's no way of preventing this kind of content from spreading like wildfire across the Internet now - so in 2025, the camera definitely lies."