The UK's elections watchdog has introduced a deepfake detection scheme to combat AI-generated misinformation ahead of the May local elections. The Electoral Commission said the system will monitor online content for AI-generated audio and video intended to mislead voters.
While deepfakes are 'yet to meaningfully affect a UK election', the commission warned the technology is becoming 'more sophisticated and more accessible'. The move follows reports that over half of voters encountered misleading information about parties or candidates during the 2024 election, and around a quarter saw or heard a deepfake.
Last year in Ireland, a deepfake falsely depicted a presidential candidate withdrawing from the race days before polling. MPs have previously cautioned that legislation must keep pace with technological developments, particularly concerning sexually explicit deepfakes targeting female candidates. Labour MP Alex Davies-Jones noted that most politicians have been targeted, and speaking out could lead to 'increased targeting'.
The pilot scheme, delivered in partnership with the Home Office, will identify deepfakes, track their impact on voter intentions, and collaborate with parties and candidates to remove misleading content. The commission will share its findings after the May elections, when voters in England elect local council representatives on 7 May.
Vijay Rangarajan, Chief Executive of the Electoral Commission, stated: 'Voters want accurate information. Deliberately misleading or abusive video of candidates must not be shared. This pilot means we can identify deepfakes quickly, track their impact, and work with parties and candidates to take down or correct misleading material.'
