2024 is one of the biggest global election years we’ve seen in decades - and it’s arriving at the same time as AI is emerging as the next technological revolution. Valent detects, tracks and responds to online manipulation, and we are already seeing how AI is posing grave threat to the electoral processes. Here’s are our key concerns:
🎯 Personalised disinformation campaigns
Disinformation campaigns tailored to individual preferences are going to become the new norm. Bad actors can deepen ideological divisions and reinforce echo chambers because they’ll know exactly how to activate your emotions and influence your decisions. This is not a problem that can be addressed by fact checkers. At the same time, as far as we know, election regulators and platforms themselves have not developed appropriate responses.
🤖 Deepfake technology and AI
Faked videos and audios claiming to represent political figures are already virtually indistinguishable from real ones due to the ascent of deepfake technology. We’re going to see everything manipulated - speeches, interviews, statements, and even fake ‘hot mic’ scenarios that will confuse voters. Even if they are quickly debunked - which will keep fact checkers busy - the cumulative impact on voters will be trust degradation.
📱 Social media manipulation
Powerful algorithms on platforms like TikTok and Instagram recommend us content based on what is performing well and our existing viewing habits. So what happens when disinformation campaigns try to take advantage of this and make viral content to game content feeds? Bad actors can use AI tools to help them craft viral content, target specific demographics, and manipulate trending topics - making talking points seem popular or widely endorsed when they aren’t.
😶🌫️ False info via “Hallucinations”
When AI churns out enormous amounts of text, it’s capable of presenting incorrect or misleading information as fact. If publishers online don’t vet AI-assisted writing, they could flood social media platforms and news websites with false claims. It’s going to be too much for fact-checkers to be able to deal with - and will provide further justification for rejecting accurate, but unpalatable, information.
So what does this mean for the 2024 elections?
🔴 Democracy and trust will be undermined - a less trusting public is a public that may lose faith in the entire electoral process, even putting them off from voting
🔴 Polarisation will deepen - AI disinformation will exacerbate existing divides and even create new ones
🔴 Digital integrity measures will come under stress - AI disinformation campaigns are going to challenge all our previous technological defences. Are existing systems up to it?
Governments, Big Tech, and civil society will need to join together to protect election integrity. We’ll be there too, helping to protect the future of democracy with our AI tool, Ariadne.
In the UK context only. We know our main political parties have some really smart staff or hire it in. How will they react to this prospect? Be it defensive or offensive. That guardian of the electoral process, the Electoral Commission - are they aware? I assume other public bodies that should be on guard are OFCOM and the Home Office, for two examples. Somehow I think not, even if they are aware of such manipulation abroad - not that the UK has never indulged!
AI is a huge money spinner for profit making companies, mostly owned outside of the UK. They should be accountable to the public. Can't we find a way to identify it, like a watermark in a digital photo?