Blog
Every US presidential election is significant, however, the 2024 race takes on heightened importance as the controversial Donald Trump faces off against Vice President Kamala Harris. Their contest has come to symbolize a struggle over the direction of the nation’s future. With so much at stake, malicious actors are increasingly using AI to manipulate election campaigns and results. In this environment, social media safeguards and heightened public awareness are more critical than ever to protect the integrity of the 2024 election.
How Bad Actors Are Using AI to Influence US Elections
1. Phishing
Before elections, candidates often reach out to voters to secure their support — a tactic cybercriminals can exploit by impersonating candidates to steal financial information or manipulate voter opinion.. One method they use is phishing.
While phishing has long been a cyber threat, earlier attempts were often riddled with errors and obvious red flags, making them easier to identify. Now, with tools like ChatGPT and other generative AI programs creating highly realistic, human-like text, phishing emails, and messages have become much more difficult to detect.
2. Deepfakes and synthetic media
Malicious actors can leverage generative AI to create fake videos, audio recordings, and photos that appear and sound altogether real. They can fabricate a video of a candidate making statements they never made or a celebrity endorsing a candidate when no such endorsement occurred. ABC News highlighted two examples: a deepfake of Taylor Swift endorsing Donald Trump and another of Ryan Reynolds wearing a Kamala Harris shirt. These deep fakes pose one of the most serious threats to the integrity of the 2024 election.
3. Data analysis and individual targeting
AI has the capability to analyze massive amounts of data, enabling it to craft personalized ad campaigns targeting specific voter demographics. It can also exploit individuals by using psychological profiles to deliver manipulative messages or spread misinformation, influencing opinions and behaviors with precision.
4. Hacking: Campaigns and election infrastructure
In August 2024, Iran hacked Trump’s campaign, offering a stark example of the dangers posed by AI-powered cyberattacks. Beyond infiltrating sensitive campaign systems and potentially accessing secure information, malicious actors can also deploy AI to target voting machines, databases, and election monitoring systems. Such attacks could lead to falsified election results, undermining the integrity of the electoral process..
5. Spreading disinformation
While disinformation may seem less dangerous than phishing or hacking, it is arguably more insidious because it fosters doubt and confusion, making it difficult for people to discern the truth . Tools like ChatGPT enable scammers to generate false information rapidly and convincingly , while AI powered bots and automation programs can disseminate it at lightning speed. By the time misinformation is debunked, the damage may already be done.
Examples of disinformation include:
- False details about Where and how to vote
- Misleading “facts” aimed at influencing voters’ opinions
- Post-election, spreading false claims regarding election results
Two Types of Cybercriminal Interference in US Elections
There are two types of bad actors when it comes to election interference:
- Cybercriminals who can exploit the high-profile nature of elections to scam unsuspecting victims for personal gain.
- Malicious actors seeking to influence election outcomes by ensuring a specific candidate wins. These are more dangerous than petty criminals, as the results of the US elections can impact the entire world.
Regardless of their motives, whether scamming or swaying election results, it is crucial for candidates, organizations, and individuals to take precautions. To avoid phishing, the rule of thumb is to never click on links from unfamiliar sources. As for disinformation, voters should remain vigilant, recognizing that not everything online is true, and always fact-check information before making decisions.