UK heading towards first ‘deep fake election’, warns new report

Less than three weeks ahead of polling day, a new report warns of the dangers of ‘realistic fake footage that is cheap to create and spread across the internet’ impacting the UK General Election.

A number of deepfake videos have appeared in recent weeks, with high-profile politicians targeted in an attempt to spread false information. In one, Wes Streeting was shown apparently calling his Labour colleague Diane Abbott a ‘silly woman’ on Politics Live. Another video falsely showed Labour’s North Durham candidate, Luke Akehurst, using crude language to mock constituents.

‘Facing Fakes’ by Matthew Feeney, Head of Tech and Innovation at the Centre for Policy Studies, warns that technological advances have made such ‘deepfakes’ easier and cheaper than ever to produce. 

However, the paper warns against the inevitable knee-jerk reaction to such technology, citing the precedent of other recent attempts to regulate new technologies.

The report recommends that the Government should focus on updating existing laws and regulations, adhering to the existing principle that it is the content itself which should be legal or illegal rather than the means of its creation. 

It urges the Government to expand on Britain’s existing work on AI safety by setting up a deepfake task force as part of its AI safety efforts, as well as sponsoring further deepfake detection contests and supporting the development of watermarking technologies.

“With three weeks before the UK heads to the polls, politicians and the public are right to be concerned about the threat deepfakes pose to democracy,” report author and CPS Head of Tech and Innovation, Matthew Feeney, said.

“However, the worst possible response would be a knee-jerk ban on the technology underpinning it.

“There is no realistic way to turn back the tide – nor should we seek to. AI-generated imagery has enormous potential in the creative industries, for example. 

“Politicians should focus on updating existing legislation, policing the content rather than the technology used to create it.

“This would mean that the Government is up to date on relevant threats and detection methods, as well as the state of deepfake detection technology.

"It would also ensure the British state is as prepared as any to tackle deepfake harms – because as the election campaign shows, we will soon be in a situation where voters may struggle to have confidence in the veracity of anything they see or hear on screen,” Feeney concluded.