With the 2024 presidential campaign already underway, the year ahead is shaping up to be unlike any previous election cycle.
The rapid growth, nearly ubiquitous use, and public interest around generative AI – alongside a shifting social media landscape and divisive political issues – will present new challenges for voters, platforms, media and marketers. In particular, the use of generative AI to create and disseminate mis- and disinformation is likely to have profound impacts.
For brands, advertising amidst these challenges comes with renewed concern around platform trust and brand safety.
How and where will AI-generated misinformation thrive? What, if anything, are the tech platforms and governments doing to help curb AI-powered misinformation? And what are the political topics likely to drive misinformation?
Generative AI and politics
The release of OpenAI’s ChatGPT has opened the floodgates for new AI tools and solutions. With options available for text, image, audio and even video, the results from AI tools are increasingly sophisticated, the use cases endless and the access limitless.
Already, Zefr has identified over 1.32 billion views of political AI content across social platforms, with political-focused misinformation content specifically increasing by 129.6% from Q1 2023 to Q2 2023.
Considering how quickly and widely content can be amplified across popular UGC-dominant social platforms, anyone in a political capacity has the ability to create and spread AI-powered content to sow confusion or distraction, build uncertainty, or persuade and manipulate.
While there is upside potential to use AI for automating processes like scaling voter outreach and fundraising faster and cheaper, the potential for campaigns, political organizations or bad actors to create convincing deepfake images, audio, video or websites as competitive tactics is deeply troubling. If unabated, it will only get more difficult for people to discern fact from fiction.
We’ve already seen instances of these information wars powered by AI around the world – from deepfake videos of Ukrainian President Volodymyr Zelensky surrendering to Russia to Pro-China bots sharing AI-generated videos of fake Chinese news anchors and outlets promoting falsehoods about the government.
Stateside, Florida Gov. Ron DeSantis’s presidential campaign earlier this year released three realistic, AI-generated false images of former President Trump with Dr. Anthony Fauci in a campaign video. In addition, there has been a substantial increase from June to July 2023 of fringe content representing 66% of overall misinformation content. This content type delves into more obscure and ominous subjects, such as illuminati conspiracies or other pseudoscientific beliefs. All of this can be used in training models as inputs for generating fake content.
The rollback of regulations
There is also growing consensus across the industry that platforms and governments aren’t doing enough to safeguard against AI-generated misinformation.
Claims of fraud in the 2020 presidential election – despite broad, substantiated debunking – remain online for voter consumption. Political accounts known to disseminate misinformation have been allowed back onto social media as well.
Government regulations have also not yet kept pace with the speed of AI development and its implications for misinformation in politics. There are currently no federal requirements for disclosing when political campaign content is AI-generated, though there is proposed legislation in Congress and by some states.
Without the guidance of disclaimers, public viewers are left to discern for themselves what is real vs. what is fake – and this is only getting more difficult. A recent University of Cambridge study found that Gen Z and millennial Americans were more susceptible to misidentifying news content as real vs. fake, and that viewers spending more time on social media in general were less likely to correctly identify AI-generated fake content vs. authentic news stories.
2024 news narratives to watch
The most recent YouGov poll for top voter issues ranked inflation, followed by health care, jobs and the economy, climate and the environment, then abortion and reproductive rights as the top five most important categories voters are following.
An analysis of viewership trends across social media for these political topics found they reflect the growing voter engagement, particularly younger generations, in increasing reliance on social media platforms as their primary news and information sources.
For brands and advertisers, coverage of these topics will continue to show up across media outlets and social platforms in both authentic and false news contexts. There is little debate that the acceleration happening in AI right now will impact the 2024 elections in both anticipated and in unexpected ways.
Foremost for advertisers, protecting audiences through verified content adjacencies and ensuring responsible digital practices against the harms of misleading election-related news content will be more important than ever.
With so much at stake, there must be a cross-industry coordinated effort, along with government regulation, when it comes to how AI is moderated and regulated. This should be of vital importance in safeguarding our democratic process and election integrity.
“Data-Driven Thinking” is written by members of the media community and contains fresh ideas on the digital revolution in media.
Follow Zefr and AdExchanger on LinkedIn.
