World must unite to combat deep fake assaults on democracy
5 min read
With a narrow lead over his main rival, Michal Šimečka was cautiously optimistic of leading his party to victory.
But two days before last year’s election in Slovakia, the internet was flooded with audio recordings of Šimečka discussing his plans to rig the vote and double the price of beer with a respected journalist.
Except he hadn’t. Both his voice and his words were the product of artificial intelligence “deepfake” technology. Under Slovakian election rules which limit what both politicians and the media can say in the 48 hours before votes are cast, Šimečka’s pro-European Progressive Slovakia (PS) party struggled to correct the falsehoods.
The precise impact on voters is unclear, but the pro-Russian nationalist party Direction – Social Democracy (Smer-SD), which has pledged to cut all aid to Ukraine and block its accession to NATO - exceeded expectations to win and now leads Slovakia’s coalition government.
The mastermind behind this blatant attack on democracy has not been publicly identified, but the prime suspect is the Kremlin, which had much to gain from a Smer-SD victory and is suspected to have form for meddling in elections.
A month after the Slovak election, Prime Minister Rishi Sunak hosted the first global summit on AI. The resulting “Bletchley Declaration”, unanimously approved by 28 nations, noted the “risks stemming from the capability to manipulate content” and added, quite rightly, that such risks “are best addressed through international cooperation”.
Since then, the UK Government has been working to develop a tool to help countries combat deepfakes and other AI technologies that could be deployed by foreign states such as Russia, Iran, and China.
The forthcoming “International Government Compact on Combating the Deceptive Use of AI by Foreign States in Elections” will set out voluntary commitments for nations to work collaboratively on research, information sharing, and improving public awareness. It could help turn the tide against political deep fakes.
Speaking at the recent Summit for Democracy in Seoul, Oliver Dowden, the Deputy Prime Minister, said several countries had already signed up – but “several” isn’t enough. Every nation that values democracy must now do its part to defend it. Every nation opposed to an AI-generated assault on freedom must now fight back.
It may not be fashionable to praise tech giants, but they have stolen a march on politicians over this crucial issue. The "Tech Accord to Combat Deceptive Use of AI in 2024 Elections", signed by 20 of the industry’s biggest players at the Munich Security Conference in February, showed a unity of purpose that governments globally would do well to emulate.
Every nation that values democracy must now do its part to defend it
And the need for collective international action has never been more urgent. Around 2 billion people in more than 60 countries will have cast their ballot by the end of this year and the deep fakers have already been busy. This year has an almost unprecedented accumulation of elections across all continents around the globe.
In India, the world’s largest democracy, sophisticated deepfake videos of A-list Bollywood stars criticising Prime Minister Narendra Modi and urging them to vote for the main opposition party in the ongoing elections have been watched more than one million times. According to a survey by the cyber security firm McAfee, 25 per cent of India’s 1.4 billion people viewed political deep fake content in the past year.
In Moldova, which borders Ukraine, deepfake footage seemed to show pro-Western President Maia Sandu endorsing a rival party sympathetic to Moscow in the run-up to local elections in January. Suspicion for that video again fell on the Kremlin.
There are mounting fears over the malevolent influence of deepfakes ahead of elections this year in the US, EU, and in the UK where, according to a survey by the Demos think tank, 57% of MPs believe they will negatively impact electoral integrity.
Ken McCullum, the Director General of MI5, has warned that AI-generated disinformation poses “a threat to democracy” and a report by GCHQ’s cyber security division said: “The UK and its allies cannot be complacent to the threat of foreign cyber-interference and attempts at influencing our democratic processes.”
Politicians including the Prime Minister Rishi Sunak and Labour leader Keir Starmer have already been subject to deep fake attacks and it is near certain that more politicians will be targeted in the run-up to the General Election. Hostile states may also feed disinformation to the electorate about where, when, and how they can vote.
AI undoubtedly offers massive opportunities, but mitigating its risks requires a strong, global response of the type that the “Compact” represents. Deep fakes don’t respect national borders, so a fractured, nation-by-nation approach is doomed to failure.
Thanks to the growing sophistication of AI technology and the ease with which it can be accessed, a deep faker in St Petersburg could today alter the words and mouth movements of a politician speaking in St Albans and disseminate it on social media within a few minutes.
When those hell-bent on undermining democracy can move that quickly, the political leaders charged with defending it must unite and respond equally fast.
Sir Brandon Lewis is a former UK security minister, secretary of state for Northern Ireland and chair of the Conservative Party.
PoliticsHome Newsletters
Get the inside track on what MPs and Peers are talking about. Sign up to The House's morning email for the latest insight and reaction from Parliamentarians, policy-makers and organisations.
