Gareth Southgate deepfake AI-generated videos 'help build morale' for England

The FA were forced to make a statement about the Gareth Southgate clips (Picture: Getty)

Deepfake AI-generated viral videos of England manager Gareth Southgate are helping to ‘build morale’ for supporters, one Three Lions fan has argued.

The Football Association (FA) are unhappy with the misleading clips that have been circulating on social media apps such as TikTok and Instagram.

Some of the videos show a fake version of Southgate making crude remarks and derogatory comments over England players both in the squad and not selected.

The problem is that many fans are fooled into thinking the clips are real – thanks to a clever AI voice generator which artificially moves Southgate’s mouth – in a setting which makes it look like the England boss is giving a real press conference.

Follow Metro.co.uk’s live blog for full coverage of England’s Euro 2024 clash vs Slovakia

‘As we do with all harmful content we will take steps to have these offensive videos removed,’ the FA said in a statement.

England fans, though, have defended the deepfake Southgate videos and have argued that they are helping to keep spirits high among supporters.

‘t’s the classic stereotypical British humour. It helps build a bit of morale to keep people motivated and happy, so I think it’s harmless fun,’ one fan told the BBC.

‘Apparently, Gareth doesn’t read anything or watch any social media anyway.’

A second supporter added that the videos were ‘quite funny’ and pointed out that Southgate is not the only person in football who has been targeted.

‘I’ve seen a lot of players and, during the season, [Jurgen] Klopp, Pep [Guardiola] and [Mikel] Arteta getting it too,’ the second fan said.

Gareth Southgate isn’t the only person in football targeted by deepfakes (Picture: Getty)

Professor Hany Farid, a digital forensics expert at the University of California, has warned people about the dangers of using deepfake AI-generated clips.

‘It’s not just people who have a big digital footprint whose identity can be co-opted,’ he said. ‘I can take as little as 30 seconds of someone’s voice, as short as a single image of somebody, and create deepfakes of them.

‘It’s one thing to have an article that is satirical, making fun of a politician or an athlete. It is very different to have a video of that person in their voice and in their likeness that can be quite offensive, and I think the courts, and we as a society, are going to have to think about where we draw that line.’