Election officials are working “rapidly” with the Home Office on a pilot project to tackle the use of deepfakes to target candidates standing in this year’s Scottish and Welsh elections.
Electoral Commission officials in Scotland said they and the Home Office hope to have software capable of detecting AI-generated deepfake videos and images up and running before the election campaign begins in late March.
Sarah McKee, head of the commission in Scotland, said that if the software detects a fake video or image, officials will contact the police, the candidate concerned and inform the public, although she acknowledged it cannot always provide 100% certainty.
He said, after this they will request the social media platform to remove the content. However, because such action is currently voluntary, the Commission also wants legally enforceable “takedown” powers that would require media platforms to remove false content.
He said the Commission had urged the UK Government to consider introducing such powers.
“What we don’t have at this time, and what we would like to have, are what’s called takedown powers — where we contact social media companies and demand that something be removed,” McKee said.
There are no known cases of deepfakes emerging during UK election campaigns, but their use has increased in elections abroad, a trend that has accelerated dramatically with the proliferation of free AI image-generation tools.
British elections and referendums have been repeatedly targeted by state-sponsored fake social media accounts From countries including Russia, Iran and North KoreaUsually designed to sow disagreement or escalate controversy.
Speaking at a pre-election briefing for journalists in Edinburgh, Mackie said the commission was also working with the Scottish Parliament and the police on a “safety and trust” project to support women and black, Asian and minority ethnic candidates who experience gender or race-based abuse or harassment.
She said a 2022 study found that almost half of all female election candidates had experienced abuse, with many saying they would not stand again. Candidates from minority-ethnic backgrounds also said the abuse intimidated them from standing again, reducing diversity at Holyrood.
Mackie said the surge in AI-powered and porn “undressing” technology, particularly generated by Elon Musk’s Grok AI platform, would fall into that category if used during the election and be reported to police.
Musk’s X and Grok platforms have been criticized for failing to remove fake, obscene and harmful content, with senior Westminster politicians demanding immediate action from the government and media regulator Ofcom.
Mackie said there was no clear legal role for the Electoral Commission or other agencies to regulate deepfakes during elections, but that both the Commission and the Home Office needed to examine what actions they could take.
He said that if this pilot project is successful then it can be started for all elections in Britain.
She said: “We don’t control election campaigns but there is a vacuum here where there are a lot of rules around the edge of the ring.
“So what we’re doing is just jumping into the center of the ring saying, OK, let’s see what we learn from this and then share it with other people.”
A Home Office spokesperson said that under the UK’s Online Safety Act, social media companies are required to remove unlawful content and prevent the spread of false information that could cause psychological or physical harm.
It added: “Protecting elections from the threat of sophisticated deepfakes is critical to maintaining public confidence in our democratic system.
“This pilot will help detect and combat deepfake content and protect the public from the influence of disinformation.”
