As malicious deepfakes continue to flood our screens with disinformation during this election year, we’ve released our 2024 Election AI Toolkit to help voters protect themselves and their vote.
Our own research reveals just how deep the problem runs. More than six in ten (63%) of Americans said they’ve seen a deepfake in the past 60 days. As for the impact of those deepfakes, nearly half (48%) who’ve seen one said it’s influenced who they’ll vote for in the upcoming election.
In all, we found that 91% of Americans said they’re concerned that AI-generated disinformation could interfere with public perception of candidates, their platforms, or even election results.
The explosive rise of disinformation, powered by AI tools
Disinformation has played a long and shady role in politics. For some time now. George Washington fell victim to it in 1777 when forged letters painted him as a British sympathizer — disinformation that followed him to the first presidency. [i]
And it’s appeared on the internet for some time too. For years, creating disinformation on the internet called for plenty of manual labor. Writers, designers, and developers all put hours into writing, creating images, and creating sites for spreading disinformation. Now, it takes just one person mere minutes. The advent of cheap and free AI tools has put disinformation into overdrive.
We’ve seen an explosive rise in malicious deepfakes in the run-up to Election Day.
- In just the first three months of 2024, the volume of deepfakes in the U.S. surged by 303% compared to the start of 2023. [ii]
- Fake news sites loaded with AI-generated photos and articles have grown from an estimated 49 sites in May 2023 to more than 700 in February 2024. [iii]
- Follow-up research pushes the estimated number of AI-powered fake news sites yet higher. In June, analysts discovered 1,265 sites targeting U.S. internet users with fake news – many posing as “local” news outlets. That figure surpasses the number of local newspapers still running in the U.S., at just 1,213 outlets. [iv]
- A British nonprofit dedicated to fighting hate speech and extremism online found that AI-generated disinformation has been rising by an average of 130% per month on X (formerly Twitter) over the past year. [v]
With polling in some states already underway, we can expect the glut of malicious deepfakes to continue. They might:
- Target politicians and their families, using deepfakes that cast them in a bad light.
- Create other deepfakes that soften the image of a candidate, known as softfakes, to make them seem more appealing.
- Spread phony polling info that prevents voters from getting to polling places in a timely way — or at all.
- Use deepfakes to skew polling results, all with the aim of influencing voters.
- Use other deepfakes to undermine confidence in the electoral process.
With that, it’s little surprise that nearly 60% of Americans say that they’re extremely or very concerned about AI’s influence on the election.[vi] Deepfakes have simply become pervasive.
Disinformation isn’t new. Using the power of AI to spread it is.
AI has given new life to the old problem of disinformation and fake news. In many ways, it’s supercharged it.
It’s done so in two primary ways:
- Bogus articles and doctored photos once took time and effort to cook up. Now, they take seconds.
- AI tools can effectively clone voices and people to create convincing-looking deepfakes in digital form.
In all, it’s easier, cheaper, and quicker than ever to create malicious deepfakes with AI tools. On top of that, the image and sound quality of deepfakes continues to improve. In all, it’s only getting tougher when it’s time to tell the difference between what’s real and what’s fake.
Taken together, this has put voters in a lurch. Who and what can they trust online?
You’re not powerless in the face of malicious AI. Quite the opposite.
Even as the creators of malicious AI-generated content have gotten cagier in their ways, their work still gives off signs of a fake. However, spotting this malicious content calls for extra effort on everyone’s part when getting their news or scrolling their feeds online. That means scrutinizing what we consume and relying on trusted fact-checking resources to get at the truth. It also means using AI as any ally, with AI tools that detect AI deepfakes in real time.
Our Election Year Toolkit will help you do just that. It covers the basics of fake news and malicious AI deepfakes, how to spot them, and more. As you’ll see, it’s a topic both broad and deep, and we explore it in a step-by-step way that helps make sense of it all for voters.
Our role in the fight against malicious deepfakes.
Sharing info about AI with voters is one of several steps we’ve taken to fight against malicious deepfakes.
In a first-of-its-kind collaboration, we’ve teamed up with Yahoo News to bolster the credibility of images on the Yahoo News platform. This collaboration integrates McAfee’s sophisticated deepfake image detection technology into Yahoo News’s content quality system, offering readers an added layer of trust.
And we’re rolling out our McAfee Deepfake detector through our partners too. It checks audio being played through your browser to figure out if the content you’re watching or listening to contains AI-generated audio. When AI audio is detected, users are notified in seconds.
Bad actors put an entirely new spin on disinformation with malicious deepfakes.
AI makes disinformation look and sound far more credible than ever. And bad actors can produce it on a tremendous scale, thanks to the ease and speed of AI tools. In an election year that calls for more scrutiny on our collective part — and our 2024 Election AI Toolkit can help. It covers how to spot a deepfake, how they spread, and several fact-checking resources that you can rely on when that bit of news you stumble across seems a little sketchy.
Toolkit Downloads
- Download the quick two-page AI Election guide here.
- Download the full McAfee AI Election Toolkit here.
References
[i] https://www.politifact.com/article/2022/feb/21/when-george-washington-fought-misinformation/
[v] https://techcrunch.com/2024/03/06/political-deepfakes-are-spreading-like-wildfire-thanks-to-genai/